CN-122025046-A - Centralized medium-volume purchasing and use management method and system for medical consumables of group type hospital based on AI prediction

CN122025046ACN 122025046 ACN122025046 ACN 122025046ACN-122025046-A

Abstract

The invention belongs to the technical field of medical consumable management, and discloses a group hospital medical consumable centralized medium purchasing and use management method and system based on AI prediction, the method comprises the steps of data acquisition and preprocessing, AI prediction model construction and training, dynamic quota calculation, real-time monitoring and comparison, intelligent early warning and feedback, data visualization and report generation, model optimization and iteration and the like. The system comprises a data acquisition module, an AI prediction modeling module, a dynamic quota management module, a real-time monitoring and comparing module, an intelligent early warning feedback module, a visual display module and a model optimization module. Through the cooperation of AI prediction, real-time data acquisition and dynamic management and control, the invention can effectively solve the problems of no use, assault use and the like of the hospital in consumable collection, realize the real-time, dynamic and accurate supervision of the protocol quantity of a plurality of hospitals and tens of thousands of consumable materials, improve the protocol quantity completion rate, reduce the management cost and ensure the clinical supply.

Inventors

ZHU YANYAN
Xue Bingtian
Deng Shaofen
CAI GUOXIN
WANG JIAPIN
CHEN JIAXUAN
YANG XIAOYAN

Assignees

广东省中医院（广州中医药大学第二附属医院、广州中医药大学第二临床医学院、广东省中医药科学院）

Dates

Publication Date: 20260512
Application Date: 20260119

Claims (10)

1. The centralized management method for the medium-volume purchase and use of medical consumables of a group hospital based on AI prediction is characterized by comprising the following steps: S1, data acquisition and preprocessing, namely acquiring historical use data, collection protocol data, clinical demand data, seasonal factor data and holiday factor data of medical consumables in each hospital area and each department of a group hospital; S2, constructing and training an AI prediction model, namely constructing an integrated prediction model integrating a plurality of machine learning algorithms based on the preprocessed data, wherein the model comprises a time sequence analysis module, a regression analysis module, a deep learning module and an expert experience adjustment module; S3, calculating the dynamic quota by utilizing a trained AI prediction model and combining factors such as total collection and acquisition protocol, time schedule, history usage rule, disease attack season and the like, and automatically calculating the annual quota and the monthly quota of each consumable; S4, real-time monitoring and comparison, namely collecting consumable usage data of each hospital area and each department in real time, and comparing the consumable usage data with a preset quota in real time; S5, intelligent early warning and feedback, namely automatically sending early warning information of different levels by the system according to the comparison result, limiting the receiving and popping up a prompt box when the usage amount exceeds the rated amount, and guiding a user to select similar substitute products; S6, data visualization and report generation, namely performing visual display on the monitoring data in a chart form, automatically generating a multi-dimensional collection and mining progress report, and supporting one-key derivation and on-line off-line feedback parallelism; and S7, model optimization and iteration, namely periodically optimizing and iterating the AI prediction model based on continuously accumulated actual use data, so as to improve the prediction accuracy.
2. The method of claim 1, wherein, The AI prediction model in the step S2 is obtained by adopting an integrated learning method, fusing an LSTM neural network, an ARIMA time sequence model, a random forest regression and a gradient lifting tree algorithm, and obtaining a final prediction result in a weighted voting mode.
3. The method of claim 1, wherein, The dynamic quota calculation in step S3 supports multiple control dimensions, including the following steps: s3-1, controlling all specification models under the same registration certificate in a unified way according to the registration certificate; S3-2, sleeve control is carried out on products checked by the sleeve, such as artificial joints, orthopedics spines and the like; s3-3, controlling according to the specification and the model, namely performing refined control on products which need to be checked to be accurate to the specification and the model.
4. The method of claim 1, wherein, And step S4, real-time monitoring and comparison are carried out by adopting a real-time data stream processing technology to acquire real-time data, so that the timeliness and accuracy of monitoring are ensured.
5. The method of claim 1, wherein, The intelligent early warning and feedback mechanism in the step S5 comprises three stages of early warning: Primary early warning and reminding, wherein when the usage amount reaches 80% of the quota, the system sends out reminding information; secondary early warning, namely when the usage amount reaches 90% of the rated amount, the system sends out warning information; And three-stage early warning and limiting, namely limiting the receiving of the system and popping up a prompt box when the usage amount exceeds the quota.
6. The method of claim 1, wherein, And step S6, data visualization and report generation support various chart types, including progress bars, trend graphs, ranking charts and early warning maps, and realize one-key viewing and one-key deriving functions.
7. The method of claim 1, wherein, The step S7 model optimization and iteration adopts an online learning algorithm, can receive new data information in real time, automatically update model parameters every month, and automatically perform model structure optimization once in a quarter.
8. The management system for centralized purchasing and using of medical consumables in a group hospital based on AI prediction is characterized by being used for executing the management method of any one of claims 1 to 7, and is of a B/S architecture, comprising a network server, a management terminal and a plurality of user terminals, wherein the network server comprises a management program, and the management program comprises: The data acquisition module is used for acquiring historical use data, collection protocol data, clinical demand data, seasonal factor data and holiday factor data of medical consumables in each hospital area and each department of the group hospital; the AI prediction modeling module is used for constructing an integrated prediction model fused with a plurality of machine learning algorithms based on the preprocessed data, and training and optimizing the model by using historical data; the dynamic quota management module is used for automatically calculating the annual quota and the monthly quota of each consumable by utilizing the trained AI prediction model and combining the factors of total collection and acquisition protocol, time schedule, history usage rule, disease attack season and the like; the real-time monitoring and comparing module is used for collecting consumable use data of each hospital area and each department in real time and comparing the consumable use data with a preset quota in real time; the intelligent early warning feedback module is used for automatically sending out early warning information of different levels according to the comparison result, and limiting the receiving and popping up the prompt box when the using amount exceeds the rated number; the visual display module is used for visually displaying the monitoring data in a chart form and automatically generating a multi-dimensional collection progress report; and the model optimization module is used for periodically optimizing and iterating the AI prediction model based on the continuously accumulated actual use data.
9. The management system of claim 8, wherein, The system also comprises a system integration interface module which is used for carrying out data butt joint and integration with the existing information systems of the hospital HIS system, the SPD system, the electronic medical record system and the hand-hemp system.
10. The management system of claim 9, wherein, The system also comprises a right management module which is used for realizing access control based on roles and supporting multi-level management right setting of a group headquarter, a hospital area, a department and the like.

Description

Centralized medium-volume purchasing and use management method and system for medical consumables of group type hospital based on AI prediction Technical Field The invention relates to the technical field of medical consumable management, in particular to a method and a system for centralized purchasing and using management of medical consumables of a group hospital based on AI prediction. Background With the deep advancement of the innovation of the medical and health system, centralized tape purchasing has become an important policy tool for controlling medical cost and optimizing resource allocation, and in recent years, the tape purchasing in high-value medical consumables is fully developed and enters a deep exploration stage. However, the floor execution of collection policies places higher demands on hospital consumable management, especially for multi-hospital, multi-level group hospitals, where the management complexity increases significantly. At present, the following technical problems exist: 1. The protocol quantity is completed, namely, the problems of no use, assault use, overuse or insufficient use and the like of report in consumable collection in hospitals are caused, so that the protocol quantity cannot be completed on time, and medical insurance examination punishment and audit examination risks are faced. 2. The data management problem is that statistics of the use data is complicated, monitoring is delayed, feedback is not timely, and real-time monitoring of consumable use conditions is difficult to achieve. 3. The management complexity is high, the group hospitals are large in scale, the departments are complete, the number of the branches is large, the consumption of consumable materials is large, and real-time, dynamic and accurate supervision is required for hundreds of hospitals and tens of thousands of consumable material protocol quantities. 4. The assessment standard is complex, the protocol quantity and the completion rate of different collection projects calculate the caliber and the assessment unit are obviously different, such as assessment according to different dimensions of registration certificate, group sleeve, specification model and the like, and the management difficulty is increased. 5. The information island problem is that the standard of each information system of a hospital is different, and the data interfaces are heterogeneous, so that the data flow from purchasing to using of consumable materials is split, and closed-loop management cannot be formed. In the prior art, the CN112908459A mainly solves the problem of how to select the lowest price supplier and can not solve the problem of agreement management and control of the tape purchase, the CN113935681A focuses on data statistics and analysis and can not provide AI prediction and dynamic management and control functions, and the CN116013545A can not solve the problem of multi-dimensional AI prediction and control of medical consumables aiming at medicine tape purchase. Therefore, a method and a system for centralized purchasing and use management of medical consumables of a group hospital based on AI prediction, convenient use and dynamic update are urgently needed to solve the technical problems. Disclosure of Invention First technical problem The invention aims to solve the technical problems of information matching problems of unnecessary consumption, assault use, excessive use or insufficient consumption of a consumable collection center report in a hospital, data aging problems of complicated usage data statistics, delayed usage data monitoring, untimely usage data feedback and the like, real-time, dynamic and accurate supervision of hundreds of hospitals, tens of thousands of consumable protocol large areas and multiple objects, management cost reduction and management efficiency improvement, intelligent decision of data driving realization, clinical supply guarantee and resource waste avoidance, and the like. (II) technical scheme In order to solve the technical problems, the invention provides a method and a system for centralized purchasing and use management of medical consumables of a group hospital based on AI prediction, which can solve the problems of a plurality of hospitals with wide area distribution in consumable collection through cooperation of software and hardware, multilevel, especially through cooperation of AI prediction, real-time data acquisition and dynamic management and control, realize real-time, dynamic and accurate supervision of hundreds of hospitals and tens of thousands of consumable protocol amounts, improve the protocol amount completion rate, reduce the management cost, ensure clinical supply and the like. The technical scheme provided by the invention is as follows: the centralized management method for the medium-volume purchase and use of medical consumables of a group hospital based on AI prediction is characterized by comprising the following steps: S1, data acqu