Artificial intelligence-based medical operation management system

Abstract

The application discloses a medical operation management system based on artificial intelligence, and relates to the technical field of operation management.The application integrates radio frequency identification, image acquisition and intelligent weighing equipment in the intelligent consumable cabinet through the internet-of-things anchoring module, realizes accurate perception and multi-dimensional data acquisition of high-value consumable warehouse-in and warehouse-out events, and generates record results based on consumable images through the number chain twin module, integrates radio frequency identification data, weight changes and operator and patient identification, constructs a data chain covering the whole life cycle of high-value consumables, and realizes consumption statistics and stock sufficiency intelligent evaluation relying on cloud data analysis.The intelligent weighing supply module generates accurate replenishment suggestions according to the daily average consumption calculated dynamically and the trend correction increment, combines the safety stock threshold, monitors and locks abnormal loss events through the weight difference rate, and triggers an alarm, so that the whole-process digital management and control of high-value consumables from warehousing, storage, use to replenishment is realized, and the management efficiency and safety of hospital consumables are effectively improved.

Description

Technical Field

[0001] This invention belongs to the field of operations management technology, specifically relating to an artificial intelligence-based medical operations management system. Background Technology

[0002] Medical institutions have placed higher demands on the full-process traceability, precise inventory control, and intelligent early warning capabilities of high-value consumables, in order to optimize resource allocation efficiency and reduce operating costs while ensuring medical quality and safety.

[0003] In existing medical operations management systems, the management of high-value consumables often relies on a combination of traditional inventory management systems and manual operations. This approach has many insurmountable drawbacks. Typically, barcode scanning or manual data entry is used to record the entry and exit of consumables, but this single-dimensional data collection method is prone to information gaps. For example, when consumables flow from the supplier to the hospital warehouse and are then used in operating rooms and other end-user locations, the handover records at intermediate stages often rely on paper documents or scattered spreadsheets, making it difficult to track the actual flow of consumables in real time. When discrepancies arise between the records and the actual inventory, managers need to spend a significant amount of time on manual inventory checks and reconciliation, which is inefficient and prone to errors. Furthermore, existing technologies lack sophisticated monitoring of consumable usage scenarios. Especially in high-value consumables usage scenarios such as surgery, the inability to strongly correlate consumable consumption with specific patients, operators, and surgical procedures creates opportunities for abnormal consumable losses, incorrect billing, and even potential medical safety hazards, making it difficult to trace the specific causes and responsible parties. Finally, existing inventory warning mechanisms are mostly based on simple upper and lower inventory limits, lacking intelligent analysis of dynamic consumption trends, resulting in delayed or inaccurate replenishment recommendations and causing inventory backlog that ties up capital. In summary, the biggest drawback of existing technologies is that their management chain is broken, information is isolated, and monitoring is passive, failing to form a complete, transparent, and traceable closed-loop management system from suppliers to patients, from warehousing to consumption.

[0004] To address the aforementioned issues, this invention proposes an artificial intelligence-based medical operations management system. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an artificial intelligence-based medical operation management system, which solves the problems of lack of full-process traceability for high-value consumables and inaccurate dynamic inventory management in existing technologies.

[0006] The objective of this invention can be achieved through the following technical solutions: An AI-based healthcare operations management system, comprising: The IoT anchoring module connects to the inventory management system, locks the high-value consumables storage area, and deploys RFID readers, image acquisition devices and intelligent weighing devices to record the entry and exit events of high-value consumables in real time, including RFID data, consumable images and weight change data. The data chain twin module generates image recording results based on consumable images; Based on RFID data, image recording results, and weight change data, combined with operator and associated patient identifiers, a complete data chain is constructed for all high-value consumables. Extract the entire process data chain of consumed high-value consumables from the cloud, perform classification and summary, generate consumption statistics reports, and perform comparative analysis with the current high-value consumables inventory in the storage area and consumption statistics reports to calculate the inventory adequacy index of various types of high-value consumables. The Smart Supply Module identifies high-value consumables to be replenished based on the stock adequacy index, assesses the average daily consumption of these consumables based on consumption statistics reports, generates replenishment suggestions by combining preset stock thresholds, dynamically updates the inventory ledger, extracts the latest generated full-process data chain of consumed high-value consumables, identifies abnormal consumable loss events, and generates alarm information.

[0007] As a further embodiment of the present invention, in the IoT anchoring module, the high-value consumable storage area is a storage area pre-set by the operator, including an intelligent consumable cabinet for storing high-value consumables, wherein high-value consumables of the same type are stored in the same intelligent consumable cabinet; The intelligent consumable cabinet is equipped with an RFID reader / writer, an image acquisition device, and an intelligent weighing device. The radio frequency identification (RFID) reader is used to interact with the RFID tags attached to the outer packaging of high-value consumables to generate RFID data. The image acquisition device is used to perform a shooting operation during radio frequency interaction to generate images of consumables; The intelligent weighing device is used to weigh the total weight of all high-value consumables in the intelligent consumable cabinet and generate weight change data.

[0008] As a further aspect of the present invention, in the IoT anchoring module, the consumable image includes the consumable image corresponding to the high-value consumable during the warehousing event and the consumable image during the outbound event. The RFID tag of the high-value consumable is bound to the unique number of the high-value consumable before the warehousing event occurs.

[0009] As a further aspect of the present invention, the specific method for generating image recording results based on consumable images in the data chain twin module is as follows: G represents any type of high-value consumables. Extract its unique identifier and denote it as B_G; Enter the high-value consumable G and its unique number B_G into the inventory database of the inventory management system, generate the storage record CG, and initialize the storage status to pending storage; Tag information M_G is generated synchronously based on the stored record CG and stored in the radio frequency tag of high-value consumable G; When a high-value consumable G is stored in the smart consumable cabinet, the RFID reader reads the tag information M_G in the RFID tag, and simultaneously retrieves the storage record corresponding to the high-value consumable G and its unique number B_G in the inventory database, verifies the storage status, and if it is to be stored, updates it to stored. The image acquisition device then takes a picture of the high-value consumable G, generates a consumable image, binds the shooting time, and stores it as the first image recording result DT1 in the storage record CG. If a high-value consumable G is stored in the intelligent consumable cabinet and its storage status is either stored or retrieved, an alarm message will be output.

[0010] As a further aspect of the present invention, in the data chain twin module, the specific method for constructing a full-process data chain linking various high-value consumables based on radio frequency identification data, image recording results, and weight change data, combined with operator and associated patient identifiers, is as follows: When a high-value consumable G is stored in an intelligent consumable cabinet, the weight data of the high-value consumable G after being stored in the intelligent consumable cabinet is obtained based on the weight change data collected by the intelligent weighing device. The weight data of the high-value consumable G before being stored in the intelligent consumable cabinet is subtracted to obtain the stored weight of the high-value consumable G, which is recorded as H1 and stored in the storage record CG. Similarly, when a high-value consumable G is taken out of the smart consumable cabinet, the weight of the high-value consumable G is determined and recorded as H2, and stored in the storage record CG. The high-value consumable G is photographed by the image acquisition device to generate a consumable image. The shooting time is bound and used as the second image recording result DT2, which is stored in the storage record CG. Synchronously verify the storage status of the storage record CG of high-value consumable G. If it is already stored, update it to be retrieved; otherwise, output an alarm message. The patient who used high-value consumable G is obtained from the hospital information management system. The patient's identifier is extracted and recorded as P. At the same time, the operator who issued high-value consumable G to patient P is extracted and recorded as DC. Both are stored in the storage record CG. Based on all the information stored in the storage record CG, arranged according to time, a full-process data chain for high-value consumables G is constructed, denoted as Link_G, and stored in the cloud.

[0011] As a further aspect of the present invention, the specific method by which the data chain twin module extracts the entire process data chain of consumed high-value consumables from the cloud, performs classification and summarization, and generates a consumption statistics report is as follows: Obtain the pre-set inventory cycle by the operator; The current moment is used as the end moment of the inventory cycle, and the start moment of the inventory cycle is determined accordingly. Extract the entire process data chain of high-value consumables consumed within the inventory cycle from the cloud; The entire process data chain is parsed, and the consumed high-value consumables are grouped according to the type of consumable. All the entire process data chains associated with the consumable type to which any high-value consumable G belongs are determined, and the entire process data chain set Link_G1, Link_G2, ..., Link_Gj is generated, where j is the total number of entire process data chains; Based on the second image recording results of j high-value consumables, determine the shooting time when j high-value consumables are taken out of the smart consumable cabinet, and integrate the j full-process data chains according to the order of shooting time to generate corresponding consumption statistics reports; Similarly, determine the consumption statistics reports for all types of consumables during the inventory period.

[0012] As a further aspect of the present invention, the specific method for calculating the stock adequacy index of various types of high-value consumables in the data chain twin module is as follows: Extract the current inventory quantity SUM of the consumable category to which high-value consumable G belongs. The current inventory quantity SUM is the high-value consumable corresponding to the consumable category in the inventory database and whose storage status is stored. Extract the average daily consumption of the corresponding consumable type for all inventory cycles except the current inventory cycle; The inventory adequacy index R_suf is calculated using R_suf=SUM / (Cou_day×T_bh), where T_bh is the preset replenishment cycle number of days. Extract the preset first stock threshold R1 and second stock threshold R2; If R_suf < R1, it is determined that the consumable type to which the high-value consumable G belongs is in a state that needs to be replenished. If R1≤R_suf<R2, it is determined that the consumable type to which the high-value consumable G belongs is in normal inventory status; If R_suf>R2, it is determined that the consumable type to which the high-value consumable G belongs is in a state of inventory backlog. Based on the replenishment status, normal inventory status, or overstock status of the high-value consumable G, generate corresponding inventory notification information.

[0013] As a further aspect of the present invention, the specific method for assessing the average daily consumption of high-value consumables to be replenished based on the consumption statistics report of the high-value consumables to be replenished in the intelligent replenishment module is as follows: Based on inventory notification information, the types of consumables that need to be replenished are identified and defined as high-value consumables awaiting replenishment. Extract the actual consumption of high-value consumables to be replenished within N inventory cycles that are adjacent to and include the current inventory cycle, perform summation and average to record as the average consumption Cou_pz, and simultaneously calculate the daily average consumption You_day for N inventory cycles, where N is a preset integer; Extract the daily consumption of any one of the N inventory cycles, and record it as a daily consumption sequence in chronological order, denoted as Q1, Q2, ..., Qm, where m is the total number of days in the inventory cycle; Construct a two-dimensional coordinate system with the timeline as the horizontal axis and the daily consumption as the vertical axis. Plot the daily consumption sequence Q1, Q2, ..., Qm as data points in the system and fit it with a straight line to lock the slope of the line. Similarly, determine the slope of the straight line associated with each of the N inventory cycles, and denote it as the straight line slope sequence K1, K2, ..., KN; The incremental factor A is calculated using the formula A = w_1 × KN + ... + w_N-1 × K2 + w_N × K1, where w_1, ..., w_N-1, and w_N are preset calculation weights, and w_1 > ... > w_N-1 > w_N, and w_1 + ... + w_N-1 + w_N = 1. The average daily consumption Qou_day in the next inventory cycle after trend correction is calculated using Qou_day = You_day + A, where Qou_day ≥ 0.

[0014] As a further aspect of the present invention, the intelligent replenishment module generates replenishment suggestions and dynamically updates the inventory ledger by combining preset upper and lower safety stock thresholds in the following specific manner: Extract the average daily consumption of high-value consumables to be replenished in the next inventory cycle (Qou_day); Extract the current inventory quantity SUM, the first inventory threshold R1, and the second inventory threshold R2 of the high-value consumables to be replenished; The total consumption Qm of high-value consumables to be replenished in the next inventory cycle is calculated using Qou_day×m. Calculate the minimum replenishment quantity X1 using R1-SUM+Qm=X1; Calculate the maximum replenishment quantity X2 using R2-SUM+Qm=X2; Generate a safe replenishment zone [X1,X2] and notify the operators; The operator determines the specific replenishment quantity and updates the inventory ledger accordingly.

[0015] As a further aspect of the present invention, the specific method by which the intelligent replenishment module locks in abnormal consumable loss events and generates alarm information is as follows: Obtain the stored weight H1 and retrieved weight H2 of any high-value consumable G, calculate the difference rate between the two, and compare it with a preset weight difference rate threshold. If the difference rate exceeds the weight difference rate threshold, it is determined that there is an abnormal loss event of high-value consumable G, and an alarm message is generated.

[0016] The beneficial effects of this invention are: This invention improves the accuracy and security of medical consumables management through intelligent methods throughout the entire process. The IoT anchoring module utilizes multi-dimensional sensing technology to capture every dynamic node of high-value consumables from warehousing to consumption in real time and without omission, ensuring the objectivity and authenticity of basic data from the source. The data chain twin module deeply integrates scattered sensing data with personnel and patient information to build an immutable end-to-end data chain, enabling precise traceability of individual items. It also identifies inventory risks in advance through automated comparative analysis and calculates the stock adequacy index to assist in scientific decision-making. The intelligent replenishment module transforms the analysis results into efficient management actions, dynamically generating accurate replenishment suggestions to optimize inventory structure and reduce operating costs. At the same time, relying on the real-time comparison capabilities of the latest data chain, it identifies abnormal consumable loss events and triggers early warnings, thereby effectively preventing management loopholes. This invention constructs a closed-loop monitoring system for high-value consumables through the collaborative operation of multi-dimensional sensing technologies. It achieves automatic and rapid identification of consumables using RFID readers, while image acquisition devices simultaneously trigger image capture during RFID interaction, providing intuitive visual evidence for every entry and exit operation and effectively solving the problem of unclear responsibility definition. Simultaneously, it introduces intelligent weighing equipment to verify the accuracy of RFID or detect unauthorized handling by monitoring subtle changes in total weight, forming a triple-verification error-proofing mechanism from electronic data, visual images, to physical weight. This improves the accuracy and real-time performance of high-value consumable inventory, reduces the manual recording burden on medical staff, and ensures the safety and traceability of high-value consumables at each stage of circulation through full-process traceability and multiple comparisons. This invention uses radio frequency identification to update the storage status in real time and trigger image acquisition, combined with weight change verification to prevent misplacement or loss of consumables. An alarm is immediately triggered in case of any abnormal status, ensuring consistency between records and actual stock. Furthermore, it automatically binds consumables to patient and operator information, constructing a complete data chain for end-to-end traceability, providing a reliable basis for medical auditing and liability determination. Based on this end-to-end data chain, it automatically generates categorized statistical reports periodically, reducing the burden of manual inventory checks. By dynamically calculating the stock adequacy index, it accurately warns of replenishment needs or inventory backlogs, assisting in the formulation of procurement plans, optimizing inventory structure, and reducing operating costs. This invention innovatively introduces time-series-based slope analysis and weighted incremental factors, assigning higher weight to recent data to capture subtle changes in consumption trends, thereby predicting the average daily demand in the next cycle. This avoids the lag and seasonal fluctuations of traditional averaging algorithms, making replenishment calculations more scientific and forward-looking. Furthermore, by combining current inventory levels with preset safety thresholds, a dynamic safe replenishment range is generated, ensuring uninterrupted clinical supply, avoiding capital tied up due to excessive reserves, and optimizing inventory cost control. Finally, by comparing weight difference rates, abnormal consumable loss events are identified and alerts are triggered, thus improving the safety and compliance of hospital consumable management while ensuring supply efficiency, achieving a dual improvement in resource efficiency and safety control. Attached Figure Description

[0017] The invention will now be further described with reference to the accompanying drawings.

[0018] Figure 1 This is a schematic diagram of the system described in this invention; Figure 2 This is a flowchart illustrating the steps described in Embodiment 3 of the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] like Figure 1 As shown, this application provides an artificial intelligence-based medical operations management system; As an embodiment 1 of this application, it specifically includes: The IoT anchoring module connects to the inventory management system, locks the high-value consumables storage area, and deploys RFID readers, image acquisition devices and intelligent weighing devices to record the entry and exit events of high-value consumables in real time, including RFID data, consumable images and weight change data. The data chain twin module generates image recording results based on consumable images; Based on RFID data, image recording results, and weight change data, combined with operator and associated patient identifiers, a complete data chain is constructed for all high-value consumables. Extract the entire process data chain of consumed high-value consumables from the cloud, perform classification and summary, generate consumption statistics reports, and perform comparative analysis with the current high-value consumables inventory in the storage area and consumption statistics reports to calculate the inventory adequacy index of various types of high-value consumables. The Smart Supply Module identifies high-value consumables to be replenished based on the stock adequacy index, assesses the average daily consumption of these consumables based on consumption statistics reports, generates replenishment suggestions by combining preset stock thresholds, dynamically updates the inventory ledger, extracts the latest generated full-process data chain of consumed high-value consumables, identifies abnormal consumable loss events, and generates alarm information.

[0021] Example 2 This embodiment further discloses the detailed steps involved in the IoT anchoring module based on Embodiment 1, specifically including the following: Firstly, this solution is mainly designed for high-value consumables. Because high-value consumables have high value and high storage requirements, they are stored in dedicated storage cabinets to ensure their quality stability.

[0022] This application needs to interface with the hospital's inventory area. First, the high-value consumables storage area is identified, which is the storage area pre-set by the operator for high-value consumables. The high-value consumables storage area includes several intelligent consumable cabinets for storing high-value consumables, and high-value consumables of the same type are stored in the same intelligent consumable cabinet. In other words, no two or more high-value consumables will appear in the same intelligent consumable cabinet.

[0023] It should also be noted that the smart consumable cabinets described in this application are all equipped with radio frequency identification readers, image acquisition devices and smart weighing devices; The radio frequency identification reader is used to interact with the radio frequency tags on the outer packaging of high-value consumables to generate radio frequency identification data associated with the corresponding high-value consumables. The image acquisition device is used to perform a shooting operation during radio frequency interaction to generate a consumable image associated with the corresponding high-value consumable; The intelligent weighing device is used to weigh the total weight of all high-value consumables in the intelligent consumable cabinet and generate weight change data.

[0024] It should be noted that the consumable images include images of high-value consumables entering and leaving the warehouse, that is, images of the high-value consumables entering and leaving the warehouse, to ensure data integrity. Before the entering event occurs, the operator binds the unique number of the high-value consumable to the RFID tag. When actual inbound and outbound events occur, i.e. when high-value consumables are placed into or removed from the cabinet, the RFID reader will emit radio waves to activate the RFID tag on the packaging of the high-value consumables, and then read the information of the high-value consumables. Through the unique number, information such as the type, price, and purpose of the high-value consumables can be directly obtained (in conjunction with the high-value consumables information database).

[0025] As described above, the multi-anchoring technology based on radio frequency identification, vision, and weight solves the pain points of discrepancies between accounts and actual stock, difficulties in inventory counting, and lack of traceability in traditional medical consumables management. It realizes a closed loop for the entire process of high-value consumables from warehousing to use, ensuring medical safety and financial security.

[0026] Example 3 This embodiment further discloses the detailed steps involved in the data chain twin module based on embodiment 2, such as... Figure 2 As shown, it specifically includes the following: First, take any type of high-value consumable as the target and perform example processing. Let the high-value consumable be G, and simultaneously obtain its unique number, which is denoted as B_G. First, an inbound operation needs to be performed on the high-value consumable G (Note: This application does not specify the specific methods for inbound and outbound operations of high-value consumables, but if the operator is responsible, prior training is required to ensure understanding of RFID and imaging operations, and to prevent interference between RFID and imaging operations. In addition, this solution supports the integration of intelligent robots for inbound and outbound operations, and ensures the normal execution of RFID and imaging operations based on the established program). First, a corresponding record, namely storage record CG, needs to be generated in the inventory database of the inventory management system based on the high-value consumable G and its unique number B_G. The storage record CG is also used to store other information, as will be explained later. At this time, the storage state of the storage record CG needs to be initialized to pending storage. At this point, the storage record CG with the storage status of "to be stored" is extracted, and the tag information M_G is generated using the storage record CG and stored in the radio frequency tag of the high-value consumable G; When a high-value consumable G is stored in a smart consumable cabinet, the first step is to use an RFID reader to read the tag information M_G in the RFID tag. After the RFID reader reads the tag information M_G, it will directly extract the high-value consumable G and its unique number B_G from the tag information M_G and upload it to the terminal (computing center). The terminal will then synchronously retrieve the storage record corresponding to the high-value consumable G and its unique number B_G from the inventory database and verify the storage status at the same time. If the verified storage status is pending storage, it is directly updated to stored and the image acquisition device is triggered to perform a shooting operation on the high-value consumable G, generate a consumable image, extract the shooting time of the shooting operation, bind it with the consumable image of the high-value consumable G, and use the binding result as the first image recording result DT1, which is stored in the storage record CG. If, during the verification process, the storage status of a CG record is either stored or retrieved, an alarm message will be output directly to prevent erroneous operations or unauthorized entry into the warehouse, thus eliminating the possibility of retrieved consumables being repeatedly put back or consumables from unknown sources being mixed in. Next, using the intelligent weighing device of the intelligent consumable cabinet, the weight difference before and after the high-value consumable G is placed in is calculated, and the stored weight H1 is obtained, specifically: When a high-value consumable G is stored in the smart consumable cabinet, the smart weighing device first collects the weight data of the high-value consumable G after it is stored in the smart consumable cabinet, and also collects the weight data of the high-value consumable G before it is stored in the smart consumable cabinet. The weight data after it is stored in the smart consumable cabinet is subtracted from the weight data before it is stored in the smart consumable cabinet to obtain the stored weight H1 of the high-value consumable G, and the stored weight H1 is stored in the storage record CG. Similarly, when a high-value consumable G is taken out of the smart consumable cabinet, the weight H2 of the high-value consumable G is determined by extracting the weight data of the high-value consumable G before it is taken out of the smart consumable cabinet and subtracting the weight data of the high-value consumable G after it is taken out of the smart consumable cabinet. Then, when the high-value consumable G is taken out of the smart consumable cabinet, an RFID tag radio frequency interaction operation is still required. At this time, the image acquisition device will be triggered to perform a shooting operation on the high-value consumable G, thereby generating a consumable image. The shooting time is extracted and bound to the consumable image to obtain the second image recording result DT2, which is stored in the storage record CG. It should be noted that after determining the stored weight H1 and the retrieved weight H2, it is also necessary to calculate the difference rate between the two and compare it with the weight difference rate threshold preset by the operator. If the difference rate exceeds the weight difference rate threshold, it is determined that there is an abnormal loss event of high-value consumable G and an alarm message is generated; otherwise, no operation is performed. The premise for determining the stored weight H1 and the retrieved weight H2 is that the operator retrieves only one high-value consumable at a time, and each high-value consumable needs to be identified by an RFID tag.

[0027] Then, continue to verify the storage status of the storage record CG of the high-value consumable G. If the storage status is "stored", then directly update the storage status to "retrieved". If the storage status is not "stored", then output an alarm message.

[0028] Next, the patient who used the high-value consumable G is retrieved from the hospital information management system by using the high-value consumable G and its unique number B_G. The patient's identifier is extracted and recorded as P. The operator (i.e., medical staff) who prescribed the high-value consumable G for the patient P is retrieved and recorded as DC. The operator and the patient P are stored in the storage record CG. Finally, based on all the information stored in the storage record CG, they are arranged in chronological order of storage time to construct the entire process data chain of high-value consumables G, denoted as Link_G, which is stored in the cloud using blockchain technology.

[0029] Next, the entire process data chain of consumed high-value consumables is extracted from the cloud, categorized and summarized, and a consumption statistics report is generated, including the following: First, obtain the pre-set inventory cycle of the operator. The duration of the inventory cycle needs to be determined in conjunction with the actual policy of the hospital, which is generally one month or one week (starting at 0:00 and 24:00 respectively). Next, we will take the current moment as the end moment of an inventory cycle as an example. By combining the duration of the inventory cycle, we can determine the start moment of the inventory cycle. Extract the entire data chain of high-value consumables that have been consumed within the current accounting cycle from the cloud (those with a storage status of "retrieved" are considered consumed); Analyze the entire process data chain of all high-value consumables, and group the consumed high-value consumables according to their type, placing high-value consumables of the same type into the same group; Here, we take any high-value consumable G and its corresponding consumable type as the target for example processing. First, we extract all the full-process data chains associated with the consumable type of the high-value consumable G, and arrange them randomly to form a set, denoted as the full-process data chain set Link_G1, Link_G2, ..., Link_Gj, where j represents the total number of full-process data chains.

[0030] Extract the second image recording results of high-value consumables corresponding to all full-process data chains in the full-process data chain set Link_G1, Link_G2, ..., Link_Gj, and extract the shooting time of the high-value consumables in the second image recording results to retrieve the intelligent consumable cabinet. Integrate the j full-process data chains according to the order of shooting time to generate the corresponding consumption statistics report. The consumption statistics report is directly displayed to the operator to assist the operator in the inventory operation. By repeating the above steps, you can obtain consumption statistics reports for all types of consumables during the inventory period.

[0031] Next, extract the consumable type to which the high-value consumable G belongs, and the current inventory quantity SUM at the current time (the current time refers to the end of the current inventory cycle). It should be noted that the current inventory quantity SUM is the high-value consumables in the inventory database that correspond to the consumable type and whose storage status is stored. Those waiting to be stored and those that have been retrieved are not included. Next, extract the average daily consumption of the corresponding consumable type in all accounting cycles (excluding the current accounting cycle) within the historical time period, and record it as Cou_day; The inventory adequacy index R_suf for this type of consumable is then calculated using the formula: R_suf=SUM / (Cou_day×T_bh), where T_bh is the number of days for the replenishment cycle preset by the operator, i.e. the time required for the supplier to deliver the goods, expressed in days. The inventory adequacy index R_suf essentially represents how long the current inventory level SUM can support consumption, divided by a replenishment cycle.

[0032] Next, extract the first stock threshold R1 and the second stock threshold R2 preset by the operator (it should be noted that the specific values ​​of the first stock threshold R1 and the second stock threshold R2 are different for different types of high-value consumables, and the operator needs to determine them in combination with the hospital's needs) and compare them with the calculated stock adequacy index R_suf. If R_suf < R1, it means that the inventory is too low and there is a risk that it will not last even one replenishment cycle. Therefore, it is necessary to replenish the inventory immediately and determine that the consumable type to which the high-value consumable G belongs is in a state that needs to be replenished. If R1≤R_suf<R2, it means that the inventory level is healthy and can be maintained until the next replenishment. Therefore, it is determined that the high-value consumable G belongs to the consumable category in normal inventory status. If R_suf > R2, it means that there is too much inventory, which ties up funds and space. An early warning is needed, and it is determined that the consumable type to which the high-value consumable G belongs is in a state of inventory backlog. Finally, based on the replenishment status, normal inventory status, or overstock status of the high-value consumable G, the corresponding inventory notification information is generated and sent to the operators.

[0033] Example 4 This embodiment, based on embodiment 3, further discloses a method for outputting turn-on and turn-off instructions based on a first-level signal and a second-level signal, specifically including the following: First, based on the inventory notification information, the types of consumables that are in the process of needing replenishment are defined as high-value consumables awaiting replenishment. Then, extract the actual consumption of high-value consumables to be replenished within N inventory cycles. It should be noted that N inventory cycles include the current inventory cycle and the N-1 inventory cycles closest to the current inventory cycle. Perform a summation and average operation on the actual consumption of N inventory cycles, and record the calculation result as the average consumption Cou_pz. Simultaneously calculate the average daily consumption You_day for N inventory cycles, where N is an integer preset by the operator.

[0034] Next, extract any one of the N inventory cycles and perform the example processing. All other inventory cycles are processed in the same and synchronous manner as the original inventory cycle. First, extract all daily consumption amounts for the inventory cycle and obtain the daily consumption sequence associated with the inventory cycle in chronological order, denoted as Q1, Q2, ..., Qm, where m is the total number of days in the inventory cycle. If the inventory cycle lasts for one month, then m = 30. Then, a two-dimensional coordinate system is constructed with the timeline as the horizontal axis and the daily consumption as the vertical axis. The daily consumption sequence Q1, Q2, ..., Qm is regarded as data points. The timeline is aligned with the horizontal axis in the two-dimensional coordinate system to obtain m data points. The data points are then fitted with a straight line, and the slope of the fitted line is calculated. By repeating the above steps, the slope of the straight line associated with each of the N inventory cycles can be determined and sorted in chronological order, denoted as the straight line slope sequence K1, K2, ..., KN, corresponding to the N inventory cycles, where KN is the current inventory cycle.

[0035] Next, the increment factor A is calculated using A=w_1×KN+...+w_N-1×K2+w_N×K1, which represents the daily average increment predicted based on the consumption trend of the recent N cycles; If the value of the increment factor A is positive, it indicates that the consumption is on the rise and the correction amount is positive. If the value of the increment factor A is negative, it indicates that the consumption is on the fall and the correction amount is negative. However, in order to ensure the correctness of the logic, when the correction amount is negative, the increment factor A is set to 0, indicating that no increment is needed. It should be noted that w_1, ..., w_N-1, and w_N are all preset calculation weights, and w_1 > ... > w_N-1 > w_N, w_1 + ... + w_N-1 + w_N = 1. The more recent the accounting period, the greater the calculation weight, and the better it reflects the recent trend.

[0036] Finally, the average daily consumption Qou_day in the next inventory cycle after trend correction is calculated by using Qou_day=You_day+A, and Qou_day is greater than or equal to 0.

[0037] Through the above operations, time series analysis and trend-weighted correction are combined, historical average consumption is considered, and the changing trend of consumption is captured. By giving higher weight to recent trends, the forecast is made closer to actual business changes, which is conducive to the dynamic adaptability of hospital business volume that may fluctuate with seasonality, sudden epidemics and other factors. Next, based on the safety stock upper and lower limits preset by the operators, replenishment suggestions are generated, and the inventory ledger is dynamically updated, as follows: First, extract the average daily consumption Qou_day of the high-value consumables to be replenished in the next inventory cycle, and simultaneously extract the current inventory SUM, the first inventory threshold R1, and the second inventory threshold R2 of the high-value consumables to be replenished. The total consumption Qm of high-value consumables to be replenished in the next inventory cycle is calculated using Qou_day×m. The minimum replenishment quantity X1 is calculated using R1-SUM+Qm=X1, and the maximum replenishment quantity X2 is calculated using R2-SUM+Qm=X2. A safe replenishment interval [X1,X2] is generated based on the minimum replenishment quantity X1 and the maximum replenishment quantity X2, and this is used as notification information to notify the operators.

[0038] Finally, the operator determines the specific replenishment quantity and updates the inventory ledger accordingly. For example, if the current inventory quantity (SUM) is 100 units and the specific replenishment quantity is 200 units, then after updating the inventory ledger, the current inventory quantity (SUM) will be 300 units.

[0039] All data in the formulas described above have been calculated with dimensions removed. Furthermore, any content not described in detail in this specification is existing technology known to those skilled in the art.

[0040] The above description is merely an example and illustration of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the invention or exceed the scope defined in the claims, all of which should fall within the protection scope of the present invention.

[0041] It should be stated that all user data collected in this application was collected with the user's consent and authorization. Furthermore, the uses of user data are legal and compliant, and the use and processing of user data comply with the relevant laws, regulations, and standards of the relevant regions.

Claims

1. A medical operation management system based on artificial intelligence, characterized in that, The system includes: The IoT anchoring module connects to the inventory management system, locks the high-value consumables storage area, and deploys RFID readers, image acquisition devices and intelligent weighing devices to record the entry and exit events of high-value consumables in real time, including RFID data, consumable images and weight change data. The data chain twin module generates image recording results based on consumable images; Based on RFID data, image recording results, and weight change data, combined with operator and associated patient identifiers, a complete data chain is constructed for all high-value consumables. Extract the entire process data chain of consumed high-value consumables from the cloud, perform classification and summary, generate consumption statistics reports, and perform comparative analysis with the current high-value consumables inventory in the storage area and consumption statistics reports to calculate the inventory adequacy index of various types of high-value consumables. The Smart Supply Module identifies high-value consumables to be replenished based on the stock adequacy index, assesses the average daily consumption of these consumables based on consumption statistics reports, generates replenishment suggestions by combining preset stock thresholds, dynamically updates the inventory ledger, extracts the latest generated full-process data chain of consumed high-value consumables, identifies abnormal consumable loss events, and generates alarm information.

2. The system according to claim 1, characterized in that, In the IoT anchoring module, the high-value consumables storage area is a storage area pre-set by the operator, including an intelligent consumables cabinet for storing high-value consumables, wherein high-value consumables of the same type are stored in the same intelligent consumables cabinet; The intelligent consumable cabinet is equipped with an RFID reader / writer, an image acquisition device, and an intelligent weighing device. The radio frequency identification (RFID) reader is used to interact with the RFID tags attached to the outer packaging of high-value consumables to generate RFID data. The image acquisition device is used to perform a shooting operation during radio frequency interaction to generate images of consumables; The intelligent weighing device is used to weigh the total weight of all high-value consumables in the intelligent consumable cabinet and generate weight change data.

3. The system according to claim 2, characterized in that, In the IoT anchoring module, the consumable image includes the consumable image of the corresponding high-value consumable during the inbound event and the consumable image during the outbound event; The RFID tag of the high-value consumable is bound to the unique number of the high-value consumable before the warehousing event occurs.

4. The system according to claim 3, characterized in that, In the aforementioned data chain twin module, the specific method for generating image recording results based on consumable images is as follows: G represents any type of high-value consumables. Extract its unique identifier and denote it as B_G; Enter the high-value consumable G and its unique number B_G into the inventory database of the inventory management system, generate the storage record CG, and initialize the storage status to pending storage; Tag information M_G is generated synchronously based on the stored record CG and stored in the radio frequency tag of high-value consumable G; When a high-value consumable G is stored in the smart consumable cabinet, the RFID reader reads the tag information M_G in the RFID tag, and simultaneously retrieves the storage record corresponding to the high-value consumable G and its unique number B_G in the inventory database, verifies the storage status, and if it is to be stored, updates it to stored. The image acquisition device then takes a picture of the high-value consumable G, generates a consumable image, binds the shooting time, and stores it as the first image recording result DT1 in the storage record CG. If a high-value consumable G is stored in the intelligent consumable cabinet and its storage status is either stored or retrieved, an alarm message will be output.

5. The system according to claim 4, characterized in that, In the aforementioned data chain twin module, the specific method for constructing a complete data chain linking various high-value consumables based on RFID data, image recording results, and weight change data, combined with operator and associated patient identifiers, is as follows: When a high-value consumable G is stored in an intelligent consumable cabinet, the weight data of the high-value consumable G after being stored in the intelligent consumable cabinet is obtained based on the weight change data collected by the intelligent weighing device. The weight data of the high-value consumable G before being stored in the intelligent consumable cabinet is subtracted to obtain the stored weight of the high-value consumable G, which is recorded as H1 and stored in the storage record CG. Similarly, when a high-value consumable G is taken out of the smart consumable cabinet, the weight of the high-value consumable G is determined and recorded as H2, and stored in the storage record CG. The high-value consumable G is photographed by the image acquisition device to generate a consumable image. The shooting time is bound and used as the second image recording result DT2, which is stored in the storage record CG. Synchronously verify the storage status of the storage record CG of high-value consumable G. If it is already stored, update it to be retrieved; otherwise, output an alarm message. The patient who used high-value consumable G is obtained from the hospital information management system. The patient's identifier is extracted and recorded as P. At the same time, the operator who issued high-value consumable G to patient P is extracted and recorded as DC. Both are stored in the storage record CG. Based on all the information stored in the storage record CG, arranged according to time, a full-process data chain for high-value consumables G is constructed, denoted as Link_G, and stored in the cloud.

6. The system according to claim 5, characterized in that, In the aforementioned data chain twin module, the specific method for extracting the entire process data chain of consumed high-value consumables from the cloud, performing classification and summarization, and generating consumption statistics reports is as follows: Obtain the pre-set inventory cycle by the operator; The current moment is used as the end moment of the inventory cycle, and the start moment of the inventory cycle is determined accordingly. Extract the entire process data chain of high-value consumables consumed within the inventory cycle from the cloud; The entire process data chain is parsed, and the consumed high-value consumables are grouped according to the type of consumable. All the entire process data chains associated with the consumable type to which any high-value consumable G belongs are determined, and the entire process data chain set Link_G1, Link_G2, ..., Link_Gj is generated, where j is the total number of entire process data chains; Based on the second image recording results of j high-value consumables, determine the shooting time when j high-value consumables are taken out of the smart consumable cabinet, and integrate the j full-process data chains according to the order of shooting time to generate corresponding consumption statistics reports; Similarly, determine the consumption statistics reports for all types of consumables during the inventory period.

7. The system according to claim 6, characterized in that, In the aforementioned data chain twin module, the specific method for calculating the stock adequacy index of various types of high-value consumables is as follows: Extract the current inventory quantity SUM of the consumable category to which high-value consumable G belongs. The current inventory quantity SUM is the high-value consumable corresponding to the consumable category in the inventory database and whose storage status is stored. Extract the average daily consumption of the corresponding consumable type for all inventory cycles except the current inventory cycle; The inventory adequacy index R_suf is calculated using R_suf=SUM / (Cou_day×T_bh), where T_bh is the preset replenishment cycle number of days. Extract the preset first stock threshold R1 and second stock threshold R2; If R_suf < R1, it is determined that the consumable type to which the high-value consumable G belongs is in a state that needs to be replenished. If R1≤R_suf<R2, it is determined that the consumable type to which the high-value consumable G belongs is in normal inventory status; If R_suf>R2, it is determined that the consumable type to which the high-value consumable G belongs is in a state of inventory backlog. Based on the replenishment status, normal inventory status, or overstock status of the high-value consumable G, generate corresponding inventory notification information.

8. The system according to claim 7, characterized in that, In the intelligent replenishment module, the specific method for assessing the average daily consumption of high-value consumables to be replenished based on the consumption statistics report is as follows: Based on inventory notification information, the types of consumables that need to be replenished are identified and defined as high-value consumables awaiting replenishment. Extract the actual consumption of high-value consumables to be replenished within N inventory cycles that are adjacent to and include the current inventory cycle, perform summation and average to record as the average consumption Cou_pz, and simultaneously calculate the daily average consumption You_day for N inventory cycles, where N is a preset integer; Extract the daily consumption of any one of the N inventory cycles, and record it as a daily consumption sequence in chronological order, denoted as Q1, Q2, ..., Qm, where m is the total number of days in the inventory cycle; Construct a two-dimensional coordinate system with the timeline as the horizontal axis and the daily consumption as the vertical axis. Plot the daily consumption sequence Q1, Q2, ..., Qm as data points in the system and fit it with a straight line to lock the slope of the line. Similarly, determine the slope of the straight line associated with each of the N inventory cycles, and denote it as the straight line slope sequence K1, K2, ..., KN; The incremental factor A is calculated using the formula A = w_1 × KN + ... + w_N-1 × K2 + w_N × K1, where w_1, ..., w_N-1, and w_N are preset calculation weights, and w_1 > ... > w_N-1 > w_N, and w_1 + ... + w_N-1 + w_N = 1. The average daily consumption Qou_day in the next inventory cycle after trend correction is calculated using Qou_day = You_day + A, where Qou_day ≥ 0.

9. The system according to claim 8, characterized in that, In the intelligent replenishment module, the specific method for generating replenishment suggestions and dynamically updating the inventory ledger by combining preset upper and lower safety stock thresholds is as follows: Extract the average daily consumption of high-value consumables to be replenished in the next inventory cycle (Qou_day); Extract the current inventory quantity SUM, the first inventory threshold R1, and the second inventory threshold R2 of the high-value consumables to be replenished; The total consumption Qm of high-value consumables to be replenished in the next inventory cycle is calculated using Qou_day×m. Calculate the minimum replenishment quantity X1 using R1-SUM+Qm=X1; Calculate the maximum replenishment quantity X2 using R2-SUM+Qm=X2; Generate a safe replenishment zone [X1,X2] and notify the operators; The operator determines the specific replenishment quantity and updates the inventory ledger accordingly.

10. The system according to claim 5, characterized in that, In the intelligent replenishment module, the specific method for identifying abnormal consumable loss events and generating alarm information is as follows: Obtain the stored weight H1 and retrieved weight H2 of any high-value consumable G, calculate the difference rate between the two, and compare it with a preset weight difference rate threshold. If the difference rate exceeds the weight difference rate threshold, it is determined that there is an abnormal loss event of high-value consumable G, and an alarm message is generated.

Images