Anesthesia medical material management system based on Internet of Things and cloud data collaboration

Abstract

The invention discloses an anesthesia medical material management system based on Internet of Things and cloud data collaboration, and the system comprises intelligent medicine vehicle terminals which are disposed in operation rooms and are used for storing, issuing and recovering anesthesia essence medicines; the cloud data collaborative management and control platform is in communication connection with the intelligent medicine vehicle terminal (100) and is used for receiving and processing the medicine operation data from the intelligent medicine vehicle terminal and generating a whole-process management instruction; and the hospital management platform is in butt joint with the cloud data collaborative management and control platform and is used for providing operation schedule information of the patient, receiving medication charging information and generating an electronic prescription. According to the invention, automatic, visual and traceable closed-loop management of the whole process of applying, storing, taking, recycling and charging traceability of the hemp essence medicine can be realized, and integrated, centralized and standardized management of consumables and airway management tools is realized, so that the management precision, the medication safety and the operating room operation efficiency are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of medical information management and intelligent medical device technology, and in particular to an anesthesia medical supplies management system based on the Internet of Things and cloud data collaboration. Background Technology

[0002] In hospital operating room operations, the management of anesthetic medical supplies (especially narcotic and psychotropic drugs) is a core aspect of ensuring medical safety and meeting regulatory requirements. However, the traditional management model is no longer adequate for the needs: it relies on manual operation and paper registration, requiring manual completion of preoperative drug preparation, intraoperative drug retrieval, postoperative residual fluid registration, and empty bottle collection. This not only easily leads to problems such as drug confusion, counting errors, and missed expiration date warnings due to fatigue and lack of concentration, but also suffers from low process efficiency. For example, preoperative drug preparation takes 15-20 minutes per surgery, emergency intraoperative drug retrieval takes 5-8 minutes, and postoperative inventory registration takes about 10 minutes. Moreover, one manager can only cover one traditional medicine cart, resulting in high labor costs. At the same time, relying on manual inventory replenishment can easily lead to stockpiling or shortages. Some top-tier hospitals waste more than 3 million yuan annually due to expired drugs and redundant inventory. Furthermore, traditional paper records are easily altered or lost, and manual verification of empty bottle recycling can create regulatory blind spots, making it difficult to meet the compliance requirements of "traceable source and destination" for anesthetic medical supplies. Moreover, the lack of data sharing between hospital information systems (HIS), surgical anesthesia management systems (AIMS), and supply chain management systems (SPD) creates information silos, making it impossible to accurately dispense medication, build a perioperative safety evidence chain, or automatically replenish supplies. Management decisions rely on experience, and traditional pharmacy carts lack the support of technologies such as the Internet of Things and AI image recognition, making it impossible to monitor key information in real time and achieve data cloud collaboration. With the increase in surgical volume and the upgrading of control requirements, there is an urgent need to transform from manual management to data-driven management through technological upgrades. Summary of the Invention

[0003] The purpose of this invention is to provide an anesthesia medical supplies management system based on the Internet of Things and cloud data collaboration. This invention enables automated, visualized, and traceable closed-loop management of narcotic and psychotropic drugs throughout the entire process from requisition, storage, retrieval, recycling to billing and traceability. It also achieves integrated, centralized, and standardized management of consumables and airway management tools, thereby significantly improving management accuracy, medication safety, and operating room operational efficiency.

[0004] The technical solution of this invention: an anesthesia medical supplies management system based on Internet of Things and cloud data collaboration, comprising: The intelligent drug cart terminal is deployed in each operating room for the storage, dispensing and recycling of narcotic and psychotropic drugs. It integrates an Internet of Things module, a weighing sensor module and an AI image recognition module to sense changes in drug inventory in real time and automatically identify drug specifications. The cloud-based data collaborative management and control platform is connected to the intelligent medicine cart terminal to receive and process medicine operation data from the intelligent medicine cart terminal and generate full-process management instructions. The hospital management platform, including the hospital information system and the surgical anesthesia management system, is connected to the cloud-based data collaborative management platform to provide patients with surgical scheduling information, receive medication billing information, and generate electronic prescriptions. The system achieves closed-loop management of narcotic and psychotropic drugs from receipt, use, recycling to billing through the coordinated operation of the intelligent medicine cart terminal, cloud data collaborative management platform and hospital management platform.

[0005] The aforementioned anesthesia medical supplies management system based on IoT and cloud data collaboration, The AI ​​image recognition module is specifically configured to perform fine-grained visual classification tasks, and its recognition process includes the following steps: First, key text regions on medicine boxes or bottles are located using object detection models; Subsequently, the text area is cropped out and fed into an optical character recognition engine for text recognition to extract the drug name, dosage specifications, and production batch number information; Ultimately, by integrating visual features with the identified text information, drugs with highly similar appearances or names are accurately classified and verified.

[0006] The aforementioned anesthesia medical supplies management system based on IoT and cloud data collaboration has constructed a multi-level security control mechanism, specifically including operation permission control and traceability and auditing; The aforementioned access control involves the intelligent pharmacy vehicle terminal being equipped with a biometric unlocking function. Anesthesiologists or nurses must authenticate their identity via fingerprint or facial recognition before they can retrieve medication. For Class I narcotic and psychotropic drugs and high-risk drugs, the system forces a dual-person verification electronic process for medication retrieval by setting drug attributes, requiring a second authorized person to verify their identity before execution.

[0007] The traceability and auditing system automatically records data and video information on the receipt, use, residual liquid registration, empty bottle recycling and destruction of each narcotic drug, forming an unalterable full traceability chain. All user operations are automatically saved to the storage server for review, auditing and data report generation.

[0008] The aforementioned anesthesia medical supplies management system based on IoT and cloud data collaboration includes the following steps for the recycling of narcotic and psychotropic drugs: the AI ​​image recognition module is used to acquire images, identify bottle types, and verify labels of empty bottles of narcotic and psychotropic drugs placed into the recycling port; at the same time, the weighing sensor module senses the weight of the recycling operation and forms a double verification with the AI ​​image recognition results; when it is detected that an empty bottle is not recycled on time, the bottle label information does not match the record, or the weight is abnormal, the system triggers an early warning mechanism in real time.

[0009] The aforementioned anesthesia medical supplies management system based on the Internet of Things and cloud data collaboration has a built-in data-driven decision-making module in the cloud data collaborative management platform. This data-driven decision-making module has the ability to analyze drug consumption, predict demand and behavior audit, and control costs. The drug consumption analysis dynamically generates visual BI reports on drug usage in each department and can identify peak drug usage such as nighttime emergency surgeries, providing data support for optimizing inventory thresholds. The demand forecasting and behavior auditing are based on big data statistical analysis of historical surgical schedules and medication records to predict future drug demand; at the same time, data such as medication retrieval response time and doctors' medication habits in the operation log are analyzed to achieve personnel behavior auditing. The cost control mentioned above is achieved by monitoring the expiration date and inventory turnover of drugs, automatically issuing early warnings for drugs nearing their expiration date, reducing waste due to expiration, and compressing redundant inventory based on consumption patterns to improve the efficiency of capital utilization.

[0010] The aforementioned anesthesia medical supplies management system based on the Internet of Things and cloud data collaboration, wherein the intelligent medicine cart terminal serves as a distributed management node, and its physical configuration supports the storage of no less than 60 drug specifications, with a total storage capacity of no less than 3,000 vials; the cloud data collaborative management and control platform is seamlessly integrated with the hospital's supply chain management system and drug traceability platform through standardized interfaces, realizing cross-system data exchange and collaborative management.

[0011] The aforementioned anesthesia medical supplies management system based on IoT and cloud data collaboration further configures the AI ​​image recognition module to perform visual analysis of behaviors during the operation process, including: Analyze the movement trajectory of empty narcotic and psychotropic drug bottles into the recycling port to identify potential illegal delivery or bottle replacement behavior; When abnormal behavior is detected, the system automatically links with nearby video surveillance equipment to capture images of the scene and generates a risk event record containing timestamps and operator information, which is then reported to the management system.

[0012] The aforementioned anesthesia medical supplies management system based on IoT and cloud data collaboration also integrates a consumables management module in the intelligent pharmacy terminal. This module achieves precise management of consumables through the following steps: Consumables warehousing sensing: For high-value consumables with RFID electronic tags, the terminal’s built-in RFID reader reads the consumables’ specifications, batches and unique identification codes in batches and quickly, and automatically completes the warehousing registration; for low-value consumables, the weight data changes obtained by the weighing sensor module assist in the verification of quantity, specifications and batches and warehousing. Consumable usage sensing and association: When high-value consumables are taken, the RFID reader detects the tag leaving the sensing area in real time, automatically triggers an inventory count event, and generates a consumption record to be confirmed associated with the current operating nurse, operating room, and patient information; the consumption of low-value consumables is determined by the amount of weight reduction sensed by the weighing sensor module. Usage Confirmation and Inventory Update: After the surgery, the nurse confirms the actual list of consumables used on the terminal interface. The system locks the data and automatically deducts inventory accordingly. When the inventory is lower than the preset threshold, the system automatically sends a replenishment reminder to the materials management department.

[0013] The aforementioned anesthesia medical supplies management system based on IoT and cloud data collaboration also integrates an airway management tool integrated management module in the intelligent medicine cart terminal. This module is used for centralized and standardized management of video laryngoscope tools, and its implementation steps are as follows: Real-time monitoring of tool status: Through the device status monitoring module integrated into the terminal, the battery level and functional integrity of the video laryngoscope are continuously or on demand monitored, and the status data is uploaded to the cloud data collaborative management platform. The data is then displayed in a visual form on the application interface to ensure the reliability of the tool. Automatic acquisition and archiving of operation image data: When using a video laryngoscope, the built-in miniature image acquisition module of the terminal automatically captures the video stream output by the device; with user authorization, key operation image segments or images are uploaded to the cloud data collaborative management platform according to preset rules or manual triggering. The cloud data collaborative management platform automatically binds the image data with the current patient information and operation timestamp, and archives it to the surgical anesthesia management system or electronic medical record to form a complete and traceable electronic record of airway operation.

[0014] The aforementioned anesthesia medical supplies management system based on IoT and cloud data collaboration includes a smart medicine cart terminal 100 comprising a first terminal for primary monitoring and a second terminal for auxiliary verification. The device status monitoring module integrated into the terminal continuously or on-demand monitors the status data of the video laryngoscope and uploads the status data to the cloud-based data collaborative management platform 200, providing visual feedback in the application interface, including: The device status monitoring module of the first terminal acquires the first status data of the video laryngoscope. If the first status data includes a preset fault code, it sends a first instruction to the cloud data collaborative management and control platform 200. After receiving the first instruction, the cloud-based data collaborative management and control platform 200 sends a verification instruction to at least one of the second terminals; The second terminal receives the verification command and establishes a connection with the first terminal; The device status monitoring module of the second terminal acquires the second status data of the video laryngoscope, detects whether the second status data contains a preset fault code, and sends the detection result to the cloud data collaborative management and control platform 200. The cloud-based data collaborative management and control platform 200 generates a visualized detection result based on at least one of the detection results, and provides feedback in a visual form on the application interface; The first terminal and at least one second terminal are connected via a wireless network. When the detection results include multiple results, a detection result is generated according to a preset weight of each second terminal. The preset weight is updated based on the historical detection result data generated by each second terminal.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention utilizes the weighing sensor and AI image recognition dual verification mechanism of the intelligent medicine cart terminal to automatically match drug labels and medical orders. Combined with biometric identification and dual-person verification functions for Class I narcotic and psychotropic drugs and high-risk drugs, it effectively avoids problems such as drug type confusion, dosage mismatch, and expired drug release that are prone to occur in traditional manual verification. It strictly meets the five-specialized management requirements and improves the safety of anesthetic medical supplies.

[0016] 2. Preoperative medication preparation is automatically pushed to the surgical schedule and medication plan through the cloud management platform, significantly shortening medication preparation time; emergency medication retrieval during surgery relies on intelligent guidance and automatic counting functions to quickly respond to medication needs; postoperative inventory registration is automatically generated by the system, eliminating the need for manual operation. At the same time, it reduces the investment of management personnel, optimizes inventory through intelligent early warning, reduces waste caused by expired drugs and redundant inventory, and improves inventory turnover efficiency.

[0017] 3. The system of this invention establishes a data link between the intelligent pharmacy vehicle terminal, the cloud-based management platform, and the Hospital Information System (HIS) and the Artificial Anesthesia Management System (AIMS), automatically recording the entire lifecycle data of narcotic and psychotropic drugs from warehousing, dispensing, use to empty bottle recycling. In the empty bottle recycling stage, AI image recognition extracts bottle information and compares it with usage records. Abnormal situations trigger real-time alerts, and all operational data is archived with timestamps, complying with the regulatory requirements for the entire process of narcotic and psychotropic drug supervision and improving audit response capabilities.

[0018] 4. The cloud-based management platform of this invention achieves seamless integration with HIS, AIMS, and SPD supply chain platforms through a multi-platform architecture, building a data collaboration system. The system can generate visualized medication reports, identify non-routine medication use, predict drug demand based on historical data, and record operator behavior data to help managers identify process bottlenecks and optimize resource allocation.

[0019] 5. This invention, through the collaborative application of RFID readers and weighing sensors, enables rapid batch warehousing of high-value consumables and accurate verification and warehousing of low-value consumables. It provides real-time monitoring of consumable usage status and automatically links information to operators, operating rooms, and patients, ensuring traceability of consumable usage. Intelligent inventory warnings and automatic replenishment reminders reduce inventory backlog and shortages, minimize waste, and improve the precision of consumable management. Simultaneously, this invention centralizes and standardizes airway management tools, monitoring device battery power and functional integrity in real time to ensure reliable tool access. It automatically collects and archives operational image data, binding it with patient information and timestamps to form a complete electronic record of airway operations, improving the standardization and data integrity of airway management. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the system principle of the present invention; Figure 2 This is a schematic diagram of the intelligent medicine cart terminal of the present invention; Figure 3 This is a schematic diagram of the recognition results from the AI ​​image recognition module. Detailed Implementation

[0021] The present invention will be further described below with reference to the accompanying drawings and embodiments, but this should not be construed as limiting the present invention.

[0022] Example: An anesthesia medical supplies management system based on IoT and cloud data collaboration. This system addresses the problems of high reliance on manual labor, low process efficiency, blind spots in supervision, and information silos in traditional anesthesia medical supplies management. It constructs a three-layer technical architecture of terminal perception, cloud collaboration, and platform linkage. By integrating technologies such as IoT, AI image recognition, and big data analysis, it achieves closed-loop management of the entire process of narcotic drugs. The specific solution is as follows: (I) System Core Architecture Design The system adopts a distributed collaborative architecture, such as Figure 1 As shown, it consists of three parts: an intelligent pharmacy vehicle terminal, a cloud-based data collaborative management and control platform, and a hospital management platform. The functions and interaction logic of each module are as follows: 1. The intelligent medication cart terminal 100, acting as a distributed sensing and operation node and a localized management unit deployed in various operating rooms, undertakes the core functions of storing, dispensing, and collecting narcotic and psychotropic drugs. It also possesses integrated management capabilities for consumables and airway management tools, such as… Figure 2 As shown, its hardware and software module configurations are as follows: IoT module: It adopts 4G / 5G+Wi-Fi dual-mode communication to upload drug operation data (drug collection time, operator, drug information) to the cloud platform in real time. It supports local storage (≥180 days of data cache) in the absence of network, and automatically re-uploads after the network is restored.

[0023] Weighing sensor module: It adopts a high-precision pressure sensor (accuracy ±0.1g) and is integrated into each medicine storage cell to monitor the changes in the weight of medicines in real time and automatically calculate the inventory (such as triggering an abnormality reminder when the weight deviation of a single sufentanil injection exceeds 0.1g).

[0024] Image recognition module: Equipped with a 2-megapixel high-definition camera (supporting 1080P shooting) and an infrared fill light component, and built-in lightweight target detection model (optimized based on YOLOv8-tiny), it realizes drug label recognition, empty bottle verification, and operation behavior monitoring. The AI ​​image recognition module is specifically configured to perform fine-grained visual classification tasks. Fine-grained visual classification (FGVC) aims to distinguish different subcategories under a large category, such as identifying drugs of different specifications. In this embodiment, the target detection model locates key text areas on the medicine box or bottle; then, the text areas are cropped and sent to the optical character recognition engine for text recognition, extracting drug name, dosage specifications, and production batch number information; finally, the visual features and the recognized text information are fused to accurately classify and verify drugs with highly similar appearances or similar names. Specifically, to address the difficulty of identifying drugs with similar appearances (such as ibuprofen 200mg and 400mg injection), a three-level recognition process of "target detection, text extraction, and feature fusion" is designed: Step 1: Locate text regions (such as dosage markings and production batch numbers) on medicine boxes / bottles using an optimized YOLOv8 model, with a detection accuracy of ≥98% and an inference speed of ≤50ms / frame.

[0025] Step 2: Use the Tesseract OCR engine (optimized with a medical text training dataset) to extract text information, supporting Chinese and English, as well as number recognition, with an accuracy rate of ≥99% (for drug label fonts).

[0026] Step 3: Integrate visual features (drug packaging color, logo position) and text features (dosage, batch number), and use the ResNet50 classification model to achieve fine-grained classification, with an accuracy rate of ≥97% for identifying drugs with similar appearances.

[0027] The final recognition result is as follows Figure 3As shown, the identification method of this invention solves the core challenge of FGVC—visual ambiguity. Although two medicine boxes may appear visually identical, the text on them (product name, dosage, manufacturer) provides a clear signal for classification. Studies have shown that this fusion of visual and textual information achieves significantly higher classification accuracy compared to purely visual methods.

[0028] Security Control Module: Equipped with a fingerprint reader (recognition rate ≥99.5%), a facial recognition camera (false recognition rate ≤0.001%), and electronic locks, supporting biometric unlocking and dual-person verification triggering mechanisms. The security control module implements access control: Based on the RBAC (Role-Based Access Control) model, it sets up three roles: physician, nurse, and administrator. For Class I controlled substances, dual-person verification (physician + nurse, each using biometrics) is required for dispensing medication. The system automatically records the verification time and personnel information. Anesthesiologists or nurses must pass fingerprint or facial authentication before dispensing medication. For Class I controlled substances and high-risk drugs, the system enforces a dual-person verification electronic dispensing process by setting drug attributes, requiring a second authorized person to pass identity verification before execution.

[0029] The core functions of the intelligent medicine cart terminal 100 include medicine storage, intelligent dispensing, and empty bottle recycling of narcotic and psychotropic drugs. The medicine storage adopts a partitioned design, supporting the storage of ≥60 drug specifications, with a total storage capacity of ≥3000 vials. Each storage compartment corresponds to a unique electronic tag (RFID), binding information such as drug name, specifications, and expiration date. Intelligent dispensing receives a surgical medication list pushed from a cloud platform. After the physician performs biometric identification, the system automatically illuminates the indicator light of the corresponding drug storage compartment. After dispensing, the weighing module automatically verifies the quantity, and AI image recognition provides secondary confirmation of the drug information. Empty bottle recycling uses an image acquisition unit and weighing sensor built into the recycling port. After an empty bottle is inserted, the system automatically identifies the bottle type, extracts the label information (drug name, batch number), and compares it with usage records. An alarm is triggered when the weight is abnormal (e.g., residual liquid in the bottle exceeds 0.5ml).

[0030] The improvement in surgical turnaround time achieved by the intelligent pharmacy cart terminal compared to traditional manual methods, as demonstrated in Table 1, is based on experiments. Link Traditional time consumption Smart medicine cart time Saving effect Preoperative medication preparation 15-20 minutes 1 minute (automatic push) ↓90% Emergency medication retrieval during surgery 5-8 minutes 15 seconds (smart guidance automatically counts) ↓96% Postoperative inventory and registration 10 minutes Automatic report generation Zero manpower Table 1 In addition, the intelligent pharmacy terminal integrates a consumables management module, which can automatically identify high-value / low-value consumables, monitor inventory in real time, and accurately link usage with patient information. The specific steps are as follows: (1) Consumables warehousing: High-value consumables: When consumables with RFID electronic tags are placed in the intelligent medicine cart terminal, the RFID reader scans them through the antenna to quickly read information such as consumable specifications, batches, and unique identifiers in batches, and uploads it to the platform processing layer for warehousing through the network transmission layer. Low-value consumables: The weight data obtained by the shelf weighing sensors in the intelligent medicine cart terminal can be used as a means of verifying the quantity, specifications, and batches.

[0031] (2) Consumables Usage and Sensing: When a nurse retrieves consumables from the cabinet, the RFID reader detects in real time that the tag has left the sensing area, triggering an inventory event. The system immediately obtains the detailed information of the consumable and, in conjunction with the nurse's employee number, the operating room information automatically associated with the system, and the patient information, generates a "pending confirmation" consumption record at the platform processing layer. For low-value consumables, the weight (quantity) reduction is detected in real time by the weighing sensor, and the usage quantity is matched to the current operator.

[0032] (3) Information Association and Confirmation: After the surgery, the nurse confirms the list of consumables actually used in the surgery through the application interface. The platform processing layer locks this final list with the patient and surgical procedure and automatically deducts inventory. When the inventory is lower than the preset threshold, the system automatically generates a replenishment reminder and sends it to the materials management department.

[0033] Furthermore, the intelligent pharmacy vehicle terminal also features an integrated management module for airway management tools. This module enables centralized and standardized configuration of airway management tools, real-time monitoring of equipment status, and automatic collection and archiving of image data. The steps are as follows: (1) Tool status monitoring: The equipment status monitoring module (such as voltage / current sensor, communication handshake circuit) integrated in the intelligent medicine cart terminal continuously or on demand monitors the power and functional integrity of tools such as video laryngoscopes. The status data is uploaded to the platform processing layer through the network transmission layer and displayed in a visual form (such as green light / red light) in the application interaction layer to ensure that the tools used each time are "available and reliable".

[0034] In a specific case, the intelligent pharmacy vehicle terminal 100 includes a first terminal for primary monitoring (e.g., the main intelligent pharmacy vehicle in the operating room) and a second terminal for auxiliary verification (a backup intelligent pharmacy vehicle in the operating room). The device status monitoring module of the first terminal acquires the first status data of the video laryngoscope (including but not limited to battery level, communication connection status, lens working parameters, fault codes, etc.). If the first status data includes a preset fault code (such as battery level below 10%, communication interruption, lens focusing abnormality), it sends a fault warning first instruction to the cloud data collaborative management platform 200. After receiving the first fault warning instruction, the cloud-based data collaborative management and control platform 200 sends a verification instruction to at least one second terminal. The instruction includes the first terminal identifier, the unique code of the video laryngoscope, and the first status data summary. After receiving the first fault warning instruction, the cloud-based data collaborative management and control platform 200 automatically selects at least one second terminal deployed in the same surgical area as the first terminal and sends a status verification instruction to it. The instruction includes the first terminal identifier, the unique code of the video laryngoscope, and the first status data summary. After receiving the status verification command, the second terminal automatically establishes an encrypted connection with the first terminal and re-collects the functional status data (i.e., the second status data) of the video laryngoscope through its own device status monitoring module. The second terminal compares the second status data with the first status data summary, detects whether it contains the same preset fault code, and uploads the detection result of "containing fault code / not containing fault code" and the second status data to the cloud data collaborative management platform 200. The cloud-based data collaborative management and control platform 200 receives the detection results. If only one second terminal provides feedback, the actual status of the video laryngoscope is determined directly based on the detection result and the first status data. If multiple second terminals provide feedback, the detection results are weighted according to the preset weights of each second terminal (the weights are dynamically updated based on historical detection accuracy, equipment calibration time, and other data) to generate the final judgment result. The final judgment result is displayed in a visual form (such as a red fault pop-up window / green normal indicator) on the management application interface and synchronized to the local display modules of the first and second terminals.

[0035] (2) Automatic image data acquisition: When using imaging equipment such as video laryngoscopes, the miniature image acquisition module (such as a video capture card) automatically captures the video stream output by the equipment. With user authorization, this module can upload video clips or images to the platform processing layer through the network transmission layer according to preset rules (such as keyframes of intubation operation) or manually triggered. The platform processing layer binds the image data with the current patient information and surgical timestamp, and automatically archives it to the AIMS or electronic medical record system to form a complete intubation operation record.

[0036] In summary, the intelligent medicine cart terminal of the present invention, through the collaborative work of the above three core modules, constructs a unified and intelligent operating room resource management platform, realizing comprehensive integrated intelligent management of medicines, consumables and airway tools, effectively improving the management efficiency, safety and data standardization level of the operating room.

[0037] 2. The cloud-based data collaborative management and control platform 200 serves as the data processing and decision-making hub. Deployed on the hospital's private cloud using a microservice architecture, it interacts with intelligent pharmacy cart terminals and the hospital management platform via standardized interfaces (RESTful APIs). Its core modules include: Data receiving and parsing module: It can process ≥100 terminal upload data per second, supports JSON format parsing, and automatically verifies data integrity (such as returning a retransmission instruction when operator information is missing).

[0038] Full-process management instruction generation module: Generates control instructions based on drug operation data, such as automatically pushing surgical schedules and medication plans to the corresponding operating room smart pharmacy before surgery, and generating billing lists after surgery and synchronizing them to the hospital management platform.

[0039] The data-driven decision-making module possesses capabilities for drug consumption analysis, demand forecasting and behavior auditing, and cost control. Drug consumption analysis utilizes the Tableau visualization engine to generate BI reports on drug usage for various departments / operating rooms (such as the top 10 monthly consumption of narcotic and psychotropic drugs, and peak nighttime emergency room drug usage times), supporting daily / weekly / monthly filtering. Demand forecasting is based on an LSTM neural network model, inputting historical surgical schedules (last 12 months) and medication records (last 24 months) to predict drug demand for the next 7 days (accuracy ≥ 92%), automatically generating replenishment suggestions. Behavior auditing analyzes operation log data (drug retrieval response time, double-verification execution rate) to generate physician medication behavior reports (e.g., marking physicians whose average drug retrieval response time exceeds 3 minutes as needing optimization). Cost control involves real-time monitoring of drug expiration dates, automatically triggering near-expiration warnings (terminal pop-up + management SMS notification) when drugs are 30 days away from expiration, and compressing redundant inventory based on inventory turnover days (target ≤ 15 days) to reduce capital occupation.

[0040] Security and Traceability Module: Utilizes blockchain technology (consortium blockchain architecture) to store the entire process data of controlled substances (receipt, use, recycling, and destruction), generating an immutable traceability chain. It supports rapid queries by drug batch number, operator, and time range. All user operations are automatically saved to the storage server for review, auditing, and data report generation. For example, each controlled substance is bound to a unique "electronic ID card" (generated from the drug batch number and production serial number). The entire process record includes: ① Receipt: operator, receipt time, drug information; ② Use: patient information, dosage, administration time; ③ Recycling: empty bottle identification result, recycling time; ④ Destruction: destruction personnel, destruction method, time. All records are synchronized to the hospital audit system and can be exported in PDF format.

[0041] 3. The Hospital Management Platform 300 serves as an information linkage and business support platform, integrating the Hospital Information System (HIS) and the Artificial Anesthesia Management System (AIMS), and enabling two-way data interaction with the cloud platform: The data output function of the hospital management platform 300 is to push patient surgery schedules (including surgery time, surgery type, and patient allergy history) and electronic prescription information (names, dosages, and usages of narcotic and psychotropic drugs prescribed by physicians) to the cloud platform. The data receiving function receives medication billing information (automatically linked to the patient's inpatient account) and drug inventory alerts (triggering a purchase request when a drug's inventory falls below a safety threshold) from the cloud platform, and generates electronic prescription archive records. Intraoperative medication information is synchronized to the AIMS system in real time and linked with the patient's vital signs data (heart rate, blood pressure) to form a perioperative medication safety evidence chain. Postoperatively, a drug consumption list is automatically generated, supporting one-click billing in the HIS system.

[0042] In this embodiment, the system ensures process implementation through a four-layer architecture: Information distribution layer: The hospital management platform serves as the source, providing core surgical and prescription data to lay the foundation for collaboration.

[0043] Command scheduling layer: After receiving information, the cloud platform transforms it into executable operation commands for the smart medicine vehicle, and sets control rules at the same time.

[0044] Terminal execution layer: The intelligent medicine cart ensures the accuracy of medication administration and recycling through a dual verification mechanism and uploads operation data in real time.

[0045] Data closed-loop layer: The cloud platform integrates data from the entire process and synchronizes it to the hospital management platform to complete billing and archiving, while triggering inventory replenishment, forming a closed loop of "distribution-execution-feedback-optimization".

[0046] In this embodiment, seamless integration with the hospital's existing systems is achieved through a standardized interface: Integration with HIS system: Adopting the HL7 FHIR protocol, it synchronizes basic patient information (name, hospital number), surgery schedule, and pushes medication billing data (automatically matching medical insurance reimbursement rules).

[0047] Integration with AIMS system: Real-time uploading of intraoperative medication records (drug name, administration time, dosage), linked with patient vital signs data (such as heart rate, blood oxygen saturation) to form a timeline visualization.

[0048] Integration with the SPD supply chain management system: When drug inventory falls below a safety threshold (e.g., when the stock of a certain drug is only twice the daily consumption), a replenishment order is automatically generated and pushed to the SPD system to achieve "replenishment on demand".

[0049] Furthermore, taking the management of anesthetic and psychotropic drugs for a patient undergoing laparoscopic surgery as an example, the system operation process is as follows: Preoperative preparation: The HIS system pushes the patient's surgical schedule (surgery time 14:00, surgical type laparoscopic cholecystectomy) and electronic prescription (1 vial of sufentanil injection and 2 vials of propofol injection) to the cloud platform; the cloud platform matches the corresponding smart medicine cart in the operating room, generates a medication list and pushes it to the terminal, and the smart medicine cart automatically lights up the indicator light of the corresponding medicine storage compartment.

[0050] Intraoperative medication retrieval: The anesthesiologist logs into the smart medication cart via fingerprint recognition. The system displays the medication list. After the physician retrieves the medication, the weighing module automatically verifies the quantity (e.g., if one vial of sufentanil injection is retrieved, the weight deviation is 0.05g, which meets the requirements). AI image recognition extracts the drug label information (batch number 20250101), and after comparing it with the prescription and finding no errors, the medication retrieval record is completed.

[0051] Postoperative recycling and billing: After the surgery, the doctor puts the empty sufentanil bottle into the recycling chute. AI image recognition confirms the bottle type and batch number (20250101), the weighing module detects the weight (the weight of the empty bottle meets the standard), and the system records the recycling information; the cloud platform generates a billing list (1 vial of sufentanil injection, unit price XX yuan), synchronizes it to the HIS system, and automatically links it to the patient's inpatient account to complete the billing.

[0052] Traceability and Audit: Administrators can query the entire process record of this sufentanil injection via the cloud platform (recipient: Dr. Zhang, receipt time 13:50; patient: Li XX, usage time 14:10; returner: Nurse Wang, return time 15:30), and can export a PDF audit report.

[0053] In summary, this invention enables automated, visualized, and traceable closed-loop management of narcotic and psychotropic drugs throughout the entire process from application, storage, dispensing, recycling to billing and traceability. It also enables integrated, centralized, and standardized management of consumables and airway management tools, thereby significantly improving management accuracy, medication safety, and operating room operational efficiency.

Claims

1. An anesthesia medical supplies management system based on Internet of Things and cloud data collaboration, characterized in that: include: The intelligent drug cart terminal (100) is deployed in each operating room for the storage, dispensing and recycling of narcotic and psychotropic drugs. It integrates an Internet of Things module, a weighing sensor module and an AI image recognition module to sense changes in drug inventory in real time and automatically identify drug specifications. The cloud-based data collaborative management and control platform (200) is connected to the intelligent medicine cart terminal (100) for receiving and processing medicine operation data from the intelligent medicine cart terminal (100) and generating full-process management instructions; The hospital management platform (300) includes a hospital information system and a surgical anesthesia management system, which are connected to the cloud data collaborative management platform (200) to provide patients with surgical scheduling information, receive medication billing information and generate electronic prescriptions; The system achieves closed-loop management of narcotic and psychotropic drugs from receipt, use, recycling to billing through the coordinated operation of the intelligent medicine cart terminal (100), cloud data collaborative management platform (200) and hospital management platform (300).

2. The anesthesia medical supplies management system based on Internet of Things and cloud data collaboration as described in claim 1, characterized in that: The AI ​​image recognition module is specifically configured to perform fine-grained visual classification tasks, and its recognition process includes the following steps: First, key text regions on medicine boxes or bottles are located using object detection models; Subsequently, the text area is cropped out and fed into an optical character recognition engine for text recognition to extract the drug name, dosage specifications, and production batch number information; Ultimately, by integrating visual features with the identified text information, drugs with highly similar appearances or names are accurately classified and verified.

3. The anesthesia medical supplies management system based on Internet of Things and cloud data collaboration as described in claim 1, characterized in that: The system establishes a multi-level security control mechanism, specifically including operation permission control and traceability and auditing; The aforementioned operation permission control means that the intelligent medicine cart terminal (100) is equipped with a biometric unlocking function. Anesthesiologists or nurses must authenticate their fingerprints or faces before they can perform the medicine retrieval operation. For Class I narcotic and psychotropic drugs and high-risk drugs, the system sets the drug attributes to force the triggering of a two-person verification electronic drug retrieval process. The second authorized person must be verified before the operation can be performed. The traceability and auditing system automatically records data and video information on the receipt, use, residual liquid registration, empty bottle recycling and destruction of each narcotic drug, forming an unalterable full traceability chain. All user operations are automatically saved to the storage server for review, auditing and data report generation.

4. The anesthesia medical supplies management system based on Internet of Things and cloud data collaboration according to claim 1, characterized in that: The recycling of the controlled substances includes the following steps: the AI ​​image recognition module is used to acquire images, identify bottle types, and verify labels of empty controlled substances bottles placed into the recycling port; at the same time, the weighing sensor module senses the weight of the recycling operation and forms a double verification with the AI ​​image recognition results; when it is detected that the empty bottle is not recycled on time, the bottle label information does not match the record, or the weight is abnormal, the system triggers an early warning mechanism in real time.

5. The anesthesia medical supplies management system based on Internet of Things and cloud data collaboration according to claim 1, characterized in that: The cloud-based data collaborative management and control platform (200) has a built-in data-driven decision-making module, which has the ability to analyze drug consumption, forecast demand and conduct behavior audits and control costs. The drug consumption analysis dynamically generates visual BI reports on drug usage in each department and can identify peak usage for non-routine drugs such as emergency surgeries at night, providing data support for optimizing inventory thresholds. The demand forecasting and behavior auditing are based on big data statistical analysis of historical surgical schedules and medication records to predict future drug demand; at the same time, data such as medication retrieval response time and doctors' medication habits in the operation log are analyzed to achieve personnel behavior auditing. The cost control mentioned above is achieved by monitoring the expiration date and inventory turnover of drugs, automatically issuing early warnings for drugs nearing their expiration date, reducing waste due to expiration, and compressing redundant inventory based on consumption patterns to improve the efficiency of capital utilization.

6. The anesthesia medical supplies management system based on Internet of Things and cloud data collaboration according to claim 1, characterized in that: The intelligent medicine cart terminal (100) serves as a distributed management node, and its physical configuration supports the storage of no less than 60 drug specifications, with a total storage capacity of no less than 3,000 vials. The cloud-based data collaborative management and control platform (200) is seamlessly integrated with the hospital's material supply chain management system and drug traceability platform through standardized interfaces to achieve cross-system data exchange and collaborative management.

7. The anesthesia medical supplies management system based on Internet of Things and cloud data collaboration according to claim 4, characterized in that: The AI ​​image recognition module is further configured to perform visual analysis of the behavior during the operation process, including: Analyze the movement trajectory of empty narcotic and psychotropic drug bottles into the recycling port to identify potential illegal delivery or bottle replacement behavior; When abnormal behavior is detected, the system automatically links with nearby video surveillance equipment to capture images of the scene and generates a risk event record containing timestamps and operator information, which is then reported to the management system.

8. The anesthesia medical supplies management system based on Internet of Things and cloud data collaboration according to claim 1, characterized in that: The intelligent medicine cart terminal (100) also integrates a consumables management module, which achieves precise management of consumables through the following steps: Consumables warehousing sensing: For high-value consumables with RFID electronic tags, the terminal’s built-in RFID reader reads the consumables’ specifications, batches and unique identification codes in batches and quickly, and automatically completes the warehousing registration; for low-value consumables, the weight data changes obtained by the weighing sensor module assist in the verification of quantity, specifications and batches and warehousing. Consumable usage sensing and association: When high-value consumables are taken, the RFID reader detects the tag leaving the sensing area in real time, automatically triggers an inventory count event, and generates a consumption record to be confirmed associated with the current operating nurse, operating room, and patient information; the consumption of low-value consumables is determined by the amount of weight reduction sensed by the weighing sensor module. Usage Confirmation and Inventory Update: After the surgery, the nurse confirms the actual list of consumables used on the terminal interface. The system locks the data and automatically deducts inventory accordingly. When the inventory is lower than the preset threshold, the system automatically sends a replenishment reminder to the materials management department.

9. The anesthesia medical supplies management system based on Internet of Things and cloud data collaboration according to claim 1, characterized in that: The intelligent medicine cart terminal (100) also integrates an integrated airway management tool management module, which is used to manage the video laryngoscope. The implementation steps are as follows: Real-time monitoring of tool status: The device status monitoring module integrated in the terminal continuously or at a preset frequency monitors the status data of the video laryngoscope and uploads the status data to the cloud data collaborative management and control platform (200), and provides feedback in a visual form on the application interface; Automatic acquisition and archiving of operation image data: When using a video laryngoscope, the built-in miniature image acquisition module of the terminal automatically captures the video stream output by the device; with user authorization, according to preset rules or manual triggering, key operation image segments or images are uploaded to the cloud data collaborative management platform (200). The cloud data collaborative management platform (200) automatically binds the image data with the current patient information and operation timestamp, and archives it to the surgical anesthesia management system or electronic medical record to form a complete and traceable electronic record of airway operation.

10. The anesthesia medical supplies management system based on Internet of Things and cloud data collaboration according to claim 9, characterized in that: The intelligent medicine cart terminal (100) includes a first terminal for primary monitoring and a second terminal for auxiliary verification; The device status monitoring module integrated into the terminal continuously or on-demand monitors the status data of the video laryngoscope and uploads the status data to the cloud-based data collaborative management platform (200), providing visual feedback in the application interface, including: The device status monitoring module of the first terminal acquires the first status data of the video laryngoscope. If the first status data includes a preset fault code, it sends a first instruction to the cloud data collaborative management and control platform (200). After receiving the first instruction, the cloud-based data collaborative management and control platform (200) sends a verification instruction to at least one of the second terminals; The second terminal receives the verification command and establishes a connection with the first terminal; The device status monitoring module of the second terminal acquires the second status data of the video laryngoscope, detects whether the second status data contains a preset fault code, and sends the detection result to the cloud data collaborative management and control platform (200). The cloud-based data collaborative management and control platform (200) generates a visual detection result based on at least one of the detection results, and provides feedback in a visual form on the application interface; The first terminal and at least one second terminal are connected via a wireless network. When the detection results include multiple results, a detection result is generated according to a preset weight of each second terminal. The preset weight is updated based on the historical detection result data generated by each second terminal.

Images