A hospitalized patient management system and method based on blockchain

Abstract

The present invention relates to the field of health care management technology, and in particular to a hospitalized patient management system and method based on blockchain. The method comprises the following steps: obtaining hospitalized patient registration data, and updating hospitalization status according to hospitalized patient registration data to obtain hospitalized patient update data; processing hospitalized resource occupancy status according to hospitalized patient update data to obtain hospitalized resource occupancy status data; dynamically predicting resource demand and optimizing allocation according to hospitalized resource occupancy status data to obtain hospitalized resource optimization allocation data; implementing resource scheduling and real-time monitoring of resource utilization according to hospitalized resource optimization allocation data to obtain hospitalized resource utilization monitoring data to be added to a preset hospitalized patient management blockchain. The present invention uses blockchain technology to respond to changes in resource demand in a timely manner, significantly improving patient management quality, operational efficiency, and resource utilization effects.

Description

Technical Field

[0001] The present invention relates to the field of health care management technology, and in particular to a hospitalized patient management system and method based on blockchain. Background Art

[0002] In modern medical institutions, inpatient management and resource scheduling are one of the core links of hospital operations, especially in large medical institutions, where the number of inpatients is huge and their needs are complex, involving the coordination and allocation of multiple resources such as beds, medical equipment, and medical staff. Traditional inpatient management systems rely on manual or semi-automatic management methods, with scattered information and lack of real-time performance. They cannot effectively respond to emergencies or rapid changes in resource supply and demand, often leading to problems such as waste of resources, insufficient supply, and long waiting times for inpatients.

[0003] In traditional hospital management, data such as patient information, resource usage, and medical staff scheduling are often scattered across multiple independent systems, lacking effective integration and sharing. The coordination efficiency between different departments is low, resulting in delayed information updates and an inability to respond to patient needs and resource changes in a timely manner. Summary of the invention

[0004] In order to solve the above technical problems, the present invention proposes a blockchain-based inpatient management system and method to solve at least one of the above technical problems.

[0005] This application provides a method for managing inpatients based on blockchain, comprising the following steps:

[0006] Step S1: Acquire inpatient registration data, and update the hospitalization status according to the inpatient registration data to obtain inpatient update data;

[0007] Step S2: Processing the inpatient resource occupancy status according to the inpatient update data to obtain inpatient resource occupancy status data;

[0008] Step S3: dynamically predict resource demand and optimize allocation based on hospital resource occupancy data to obtain hospital resource optimization allocation data;

[0009] Step S4: Implement resource scheduling and monitor resource utilization in real time according to the inpatient resource optimization allocation data to obtain inpatient resource utilization monitoring data to add to the preset inpatient management blockchain.

[0010] In the present invention, by using blockchain technology, the data generated at each step, including patient registration data, resource occupancy data, resource optimization allocation data, etc., will be recorded in the blockchain, so that all participants (such as hospital managers, medical staff) can view the data in real time, and the data is tamper-proof and transparent to prevent malicious tampering and information leakage. The system can dynamically predict and optimize the use of hospital resources (such as beds, equipment, medical staff, etc.) to avoid idle or overused resources. Resource prediction is carried out through intelligent algorithms to ensure that the hospital can reasonably plan its resources in a certain period of time in the future, reduce unnecessary waste and improve overall efficiency. The actual use of resources is tracked through the Internet of Things (IoT) and real-time monitoring technology. The system can provide real-time feedback on the use status of beds, equipment and personnel, and dynamically adjust according to actual conditions. All these data are written to the blockchain in real time to ensure the transparency and tamper-proof of the monitoring data. The present invention can significantly improve the patient's experience during hospitalization. Through more effective resource allocation, patients can receive timely care, sufficient resources and efficient medical services during hospitalization, reducing waiting time and unnecessary transfers.

[0011] Preferably, step S1 specifically includes:

[0012] Step S11: Obtain patient identity data, and perform electronic file query based on the patient identity data to obtain the patient's electronic file data;

[0013] Step S12: Generate a unique patient identification according to the patient's electronic file data to obtain the patient's unique identification data;

[0014] Step S13: Perform on-chain registration records based on the patient's unique identification data to obtain inpatient registration data;

[0015] Step S14: generating initial hospitalization data according to the inpatient registration data to obtain initial hospitalization data;

[0016] Step S15: Perform status update confirmation based on the initial hospitalization status data to obtain updated inpatient data.

[0017] The present invention ensures the accuracy and completeness of patient information by obtaining the patient's identity data and combining it with electronic file query. All information input and records are verified by the system, reducing errors caused by manual input. The system automatically extracts the patient's historical data from the electronic file, reducing the workload of manual search and integration, shortening the time for patient information query and retrieval, and ensuring that medical staff can quickly obtain the patient's historical information, thereby improving work efficiency and data utilization. The unique identifier (UPI) generated based on the patient's electronic file ensures the independence and uniqueness of each patient in the system. All the patient's data, medical records and related operations are bound to the unique identifier, avoiding data confusion or overlap, and ensuring the continuity and consistency of the patient's data in different systems and processes. Writing the patient registration data into the blockchain ensures the transparency and immutability of the data. The distributed ledger characteristics of blockchain technology enable patient registration data to be synchronized and shared in different nodes, thereby preventing single point failures and data loss. Through the patient's unique identifier and registration record, the system automatically generates initial hospitalization data, including ward allocation, initial resource occupancy, etc. Through the initial hospitalization data, the system can automatically confirm the patient's hospitalization status and update it on the blockchain. The real-time status update ensures that the patient's hospitalization information is always up to date, which provides accurate decision-making support for medical staff and hospital managers.

[0018] Preferably, step S2 specifically includes:

[0019] Step S21: Obtaining hospitalization resource occupancy data;

[0020] Step S22: Analyze the inpatient resource occupancy situation according to the inpatient resource occupancy data and the inpatient update data to obtain inpatient resource occupancy analysis data;

[0021] Step S23: Record the resources on the chain according to the inpatient resource occupancy analysis data to obtain inpatient resource occupancy data.

[0022] In the present invention, the occupancy data of various resources (such as beds, equipment, medical staff, etc.) in the hospital are collected in real time through the Internet of Things (IoT) devices, and the latest resource usage is obtained at any time to ensure the reasonable allocation of resources and rapid response to patient needs. By combining the real-time inpatient resource occupancy data with the inpatient update data, the system can automatically analyze the actual occupancy of resources. Including which resources are being used, which resources are idle, the frequency and time period of use, etc. According to the resource occupancy analysis data, the hospital can adjust and optimize resource scheduling in a timely manner. For high-occupancy resources, the pressure can be relieved by deploying idle resources to prevent resource overload; for idle resources, they can be fully utilized by reallocation. The analysis data of inpatient resource occupancy is written into the blockchain to ensure the transparency and immutability of resource usage records. Blockchain technology ensures that all changes in resource occupancy are clearly recorded, and past resource usage can be checked at any time, and the authenticity and compliance of the data are guaranteed. Resource occupancy analysis data provides a key reference for subsequent resource demand forecasting. By analyzing past resource usage patterns, hospitals can better predict future resource needs, such as how the occupancy of beds and equipment changes in specific seasons or time periods. This information provides strong data support for hospitals to plan ahead and avoid resource shortages during peak demand periods.

[0023] Preferably, step S3 is specifically:

[0024] Step S31: extracting characteristics of the inpatient resource occupancy situation according to the inpatient resource occupancy situation data to obtain inpatient resource occupancy situation characteristic data;

[0025] Step S32: forecasting the demand for inpatient resources based on the characteristic data of inpatient resource occupancy to obtain forecast data for inpatient resource demand;

[0026] Step S33: constructing an inpatient resource optimization allocation model according to the inpatient resource demand forecast data to obtain an inpatient resource optimization resource allocation model;

[0027] Step S34: generating an inpatient resource optimization plan according to the inpatient resource optimization resource allocation model to obtain inpatient resource optimization plan data;

[0028] Step S35: Perform on-chain allocation records according to the inpatient resource optimization plan data to obtain inpatient resource optimization allocation data.

[0029] In the present invention, by extracting features from the inpatient resource occupancy data, key patterns and trends in resource use can be identified, including the time dimension, department dimension, and frequency of use of wards and equipment. The system can predict resource demand for a period of time in the future based on the extracted resource occupancy feature data, which helps hospitals prepare resources in advance and prevent resource shortages during the peak period of patients. The prediction model can adjust parameters according to historical data to make the prediction more accurate. The resource optimization allocation model combines resource demand prediction data and is constructed through optimization algorithms (such as linear programming, genetic algorithms, etc.). It can calculate the optimal resource allocation plan based on a variety of constraints (such as the number of patients, department resource conditions, etc.). This ensures efficient use of resources and reduces unnecessary waste. According to the optimization allocation model, the system automatically generates a resource optimization plan to ensure that resource allocation is flexible and adaptable, and can dynamically respond to changes in hospital resource demand and automatically adjust the allocation strategy. Whether it is short-term sudden demand or long-term resource planning, the system can make reasonable decisions based on data. All resource optimization plan data will be recorded on the blockchain to ensure the transparency and traceability of the allocation process, which not only ensures the fairness of resource allocation, but also provides a powerful supervision tool for the hospital's management and external audit. The immutable nature of blockchain ensures that all resource allocation decisions can be traced and verified.

[0030] Preferably, the feature extraction of hospitalization resource occupancy is specifically as follows:

[0031] Perform data repair according to the inpatient resource occupancy data to obtain inpatient resource occupancy repair data;

[0032] Perform target coding according to the inpatient resource occupancy repair data to obtain inpatient resource occupancy target coding data;

[0033] Performing hierarchical feature extraction based on the target coding data of inpatient resource occupancy to obtain the first inpatient resource occupancy characteristic data;

[0034] Cluster calculation is performed based on the target coding data of hospitalization resource occupancy to obtain cluster data of hospitalization resource occupancy;

[0035] Extract cluster distribution features of the cluster data of hospitalization resource occupancy to obtain characteristic data of the second hospitalization resource occupancy situation;

[0036] Perform characteristic density calculation according to the first inpatient resource occupancy characteristic data and the second inpatient resource occupancy characteristic data, to obtain the first inpatient resource occupancy characteristic density data and the second inpatient resource occupancy characteristic density data respectively;

[0037] Performing feature mutual information analysis based on the first inpatient resource occupancy feature data and the second inpatient resource occupancy feature data to obtain inpatient resource occupancy feature mutual information data;

[0038] According to the first inpatient resource occupancy characteristic density data, the second inpatient resource occupancy characteristic density data and the inpatient resource occupancy characteristic mutual information data, nonlinear feature fusion is performed on the first inpatient resource occupancy characteristic data and the second inpatient resource occupancy characteristic data to obtain the inpatient resource occupancy characteristic data.

[0039] In the present invention, through the data repair step, the system can handle missing values ​​or abnormal values ​​in the inpatient resource occupancy data to ensure the integrity and accuracy of the data. The target encoding step converts the categorical variables into numerical data suitable for model calculation, which allows complex non-numerical data to participate in model construction and feature extraction, enhancing the data processing capability of the system. Through the hierarchical feature extraction step, the system can extract multiple levels of features from the inpatient resource occupancy data. The features of different resource types, departments, and time dimensions can be analyzed in layers to ensure a fine-grained understanding of the occupancy of various types of resources. Through cluster analysis, the system can group the inpatient resource occupancy data according to similarity, identify the patterns and characteristics under different resource usage scenarios, help discover the resource occupancy patterns of specific departments or wards, and provide deep insights for optimizing allocation. Cluster distribution feature extraction helps the system identify the distribution of resources in each cluster, revealing the differences in resource use in different clusters, allowing hospitals to find out which resources are overused or which areas are under-resourced, facilitating more effective resource scheduling. Through feature density calculation, the system can evaluate the density and sparsity of the extracted resource occupancy feature data, which helps to determine the importance of certain specific features and helps the model distinguish the key features in different resource usage scenarios. Mutual information analysis reveals the correlation between different resource occupancy features. By quantifying the interdependence between features. Nonlinear feature fusion improves the feature expression ability and ensures that the resource occupancy pattern can be accurately captured and described by the fused features, thereby improving the accuracy of the resource prediction and allocation model.

[0040] Preferably, the characteristic density is calculated as follows:

[0041] Estimating the distribution type according to the first inpatient resource occupancy characteristic data and the second inpatient resource occupancy characteristic data, to obtain first characteristic distribution type estimation data and second characteristic distribution type estimation data respectively;

[0042] Using the first characteristic distribution type estimation data, the first inpatient resource occupancy characteristic data is labeled with a distribution type to obtain first characteristic distribution type labeled data, and using the second characteristic distribution type estimation data, the second inpatient resource occupancy characteristic data is labeled with a distribution type to obtain second characteristic distribution type labeled data;

[0043] Perform kernel density estimation and distribution fitting based on the first feature distribution type labeled data and the second feature distribution type labeled data to obtain first feature distribution density data and second feature distribution density data;

[0044] Density calculation is performed based on the first characteristic distribution density data and the second characteristic distribution density data to obtain first inpatient resource occupancy characteristic density data and second inpatient resource occupancy characteristic density data, respectively.

[0045] The distribution type estimation in the present invention helps to enhance the system's grasp of resource occupancy characteristics, reduce erroneous analysis caused by misjudgment of data distribution type, and improve the prediction performance of the model. The first and second feature data are marked with distribution types, so that the system can standardize the distribution of different feature data. By marking the distribution type of each feature, the system can use appropriate statistical methods in the subsequent density estimation process to make the calculation of feature density more consistent and accurate. Kernel density estimation is a hypothesis-free distribution fitting method. By performing kernel density estimation on feature data, the system can more flexibly fit the actual distribution of resource occupancy characteristics, so that density estimation can accurately reflect the local characteristics and changing trends of the data without being restricted by the preset distribution model. Distribution fitting enhances the system's understanding of resource usage patterns and helps to improve the performance of the overall prediction and optimization process. By performing density calculation on feature distribution density data, the system can quantify the density and sparsity of features. The higher the feature density, the higher the influence or frequency of the feature in the resource occupancy process; features with lower density may represent rarer resource usage.

[0046] Preferably, the nonlinear feature fusion is specifically as follows:

[0047] A density-weighted feature mutual information matrix is ​​generated according to the first inpatient resource occupancy characteristic density data, the second inpatient resource occupancy characteristic density data, and the inpatient resource occupancy characteristic mutual information data to obtain density-weighted feature mutual information matrix data;

[0048] Constructing a feature interaction network for the first inpatient resource occupancy feature data and the second inpatient resource occupancy feature data according to the density-weighted feature mutual information matrix data to obtain feature interaction network data;

[0049] A regression model and a neural network are constructed according to the feature interaction network data to obtain a first nonlinear interaction feature model and a second nonlinear interaction feature model respectively;

[0050] Generate nonlinear interaction features according to the first nonlinear interaction feature model and the second nonlinear interaction feature model to obtain first nonlinear interaction feature data and second nonlinear interaction feature data respectively;

[0051] Performing nonlinear feature fusion on the first inpatient resource occupancy condition feature data and the second inpatient resource occupancy condition feature data according to the first nonlinear interactive feature data to obtain first occupancy condition nonlinear feature data;

[0052] According to the second nonlinear interactive feature data, nonlinear feature fusion is performed on the first inpatient resource occupancy feature data and the second inpatient resource occupancy feature data to obtain second occupancy nonlinear feature data;

[0053] The first occupancy situation nonlinear characteristic data and the second occupancy situation nonlinear characteristic data are arranged in layers to obtain the inpatient resource occupancy situation characteristic data.

[0054] In the present invention, through the generation of density-weighted feature mutual information matrix, the system can extract deeper nonlinear relationships from resource occupancy features, can capture the complex interdependence and effects between features, so that the model can not only understand the independent influence of each feature, but also identify the interactive influence between them. The improvement of feature expression ability helps to improve the predictive ability of the model, especially when processing complex inpatient resource data, it can better reveal hidden patterns and associations and improve the overall performance of the system. The construction steps of regression model and neural network provide tailored processing methods for different types of features. Regression model can capture linear relationships in resource occupancy, while neural network is good at processing high-dimensional and nonlinear data. By combining these two models, the system can handle more diverse features and can generate more expressive nonlinear interactive features. The generation of nonlinear interactive features improves the generalization ability of the model, so that the system can more effectively cope with the changing situation in the use of inpatient resources, and can adapt to changes in uncertainty and complexity. Feature fusion enables the system to make full use of the interaction between multiple features, improves the data processing ability, and ensures that the system's performance in resource demand prediction and allocation optimization is more comprehensive and accurate. The hierarchical processing method can retain the differences between different features, while unifying the output format and improving the systematic nature of the processing.

[0055] Preferably, the construction of the inpatient resource optimization allocation model is specifically as follows:

[0056] Calculate the difference between supply and demand of inpatient resources based on the forecast data of inpatient resource demand, and obtain the difference between supply and demand of inpatient resources;

[0057] According to the difference data of hospitalization resource supply and demand, the hospitalization resource cost minimization constraint is performed to obtain the hospitalization resource supply optimization data;

[0058] Prioritize hospitalization resources according to the hospitalization resource supply optimization data to obtain hospitalization resource priority ranking data;

[0059] Based on the priority ranking data of inpatient resources, simulated allocation and iterative optimization are carried out to build an inpatient resource optimization resource allocation model.

[0060] In the present invention, by calculating the supply and demand difference of the forecast data of inpatient resource demand, the system can identify the gap between resource supply and demand, ensuring that the hospital can adjust resource allocation in time. The precise quantification of the supply and demand difference enables the system to determine which resources are in oversupply and which resources are in short supply, thereby laying the foundation for optimization. By constraining the cost minimization of the inpatient resource supply and demand difference data, the system can minimize the hospital operating costs while optimizing resource supply, ensuring that the needs of inpatients are met at the lowest cost, thereby improving the economic benefits of the hospital. Resource priority sorting is performed according to the inpatient resource supply optimization data to ensure that the hospital can allocate resources according to the scarcity and importance of resources. The priority sorting process helps the hospital make quick decisions, give priority to meeting the most important or most urgent resource needs, and ensure that key resources will not be wasted or unevenly distributed due to improper supply. Through simulated allocation and iterative optimization, the system can test and adjust different resource allocation strategies in a virtual environment. Each iteration can continuously optimize the resource allocation strategy, thereby gradually approaching the optimal allocation plan, reducing the uncertainty in decision-making, and ensuring that resources are optimally allocated in practical applications. Through supply and demand difference calculation, cost minimization and priority sorting, resources can be dynamically adjusted to meet actual needs and avoid idle or overused resources. Simulation allocation and iterative optimization further improve the precise allocation of resources, ensuring that hospitals use resources efficiently while avoiding waste. The system can adjust allocation strategies in real time according to changes in resource demand (such as the number of patients, changes in conditions, etc.). By continuously monitoring the difference between supply and demand and making dynamic adjustments, the system can respond quickly when resource demand changes, ensuring the flexibility and adaptability of resource allocation.

[0061] Preferably, step S4 is specifically:

[0062] Step S41: generating an inpatient resource scheduling plan according to the inpatient resource optimization allocation data to obtain inpatient resource scheduling plan data;

[0063] Step S42: collecting and adjusting IoT data according to the inpatient resource scheduling plan data to obtain inpatient resource scheduling collection data;

[0064] Step S43: Collect data based on the inpatient resource scheduling and perform on-chain monitoring and recording to obtain inpatient resource utilization monitoring data.

[0065] The precise resource scheduling plan in the present invention improves the accuracy of resource allocation, avoids over-allocation or under-allocation in resource allocation, and ensures that the hospital's resources are used most effectively. Through real-time monitoring and collection of resources by Internet of Things (IoT) devices, the system can keep track of resource usage at any time. Including ward occupancy rate, equipment usage status, workload of medical staff, etc., real-time data can be fed back to the system to help hospitals quickly make resource scheduling adjustments when demand or occupancy changes. According to the real-time resource occupancy information collected by the IoT data, the system can automatically adjust the resource scheduling plan. When a certain type of resource is in short supply or redundant, the system can automatically reallocate or optimize scheduling to ensure that resources are always kept in the best state of use. All operations, adjustments and monitoring of inpatient resource scheduling are recorded on the blockchain to ensure that each step of resource use is transparent and tamper-proof. Blockchain technology ensures the traceability of all resource scheduling decisions, and any abnormality or improper resource use can be traced back to find the cause. Through on-chain monitoring records, inpatient resource scheduling data can be securely stored and managed. The distributed nature and encryption protection of blockchain ensure the reliability of resource scheduling data and prevent data loss or tampering due to human or system failures.

[0066] Preferably, the present application also provides a blockchain-based inpatient management system for executing the blockchain-based inpatient management method as described above, the blockchain-based inpatient management system comprising:

[0067] The hospitalization status update module is used to obtain the inpatient registration data, and update the hospitalization status according to the inpatient registration data to obtain the inpatient update data;

[0068] The inpatient resource occupancy processing module is used to process the inpatient resource occupancy according to the inpatient update data to obtain the inpatient resource occupancy data;

[0069] The inpatient resource optimization allocation module is used to dynamically predict resource demand and optimize allocation based on the inpatient resource occupancy data to obtain inpatient resource optimization allocation data;

[0070] The inpatient resource utilization monitoring module is used to implement resource scheduling and monitor resource utilization in real time based on the inpatient resource optimization allocation data, and obtain inpatient resource utilization monitoring data to add to the preset inpatient management blockchain.

[0071] The beneficial effect of the present invention is that by closely combining the updated data of inpatients with the resource occupancy, and through dynamic resource demand prediction and optimized allocation, the present invention realizes highly intelligent management of inpatient resources. The system can accurately predict the resource demand for a period of time in the future according to the number of patients, changes in the condition, resource occupancy history, etc., reduce the imbalance between supply and demand of resources, and avoid resource shortages or waste caused by emergencies or surges in patients. Blockchain technology is used to manage all key data related to inpatients, including inpatient registration data, resource occupancy data, resource allocation optimization data, and real-time monitoring data. The immutability of blockchain ensures that all information can be traced and verified, prevents management errors caused by data modification or human operation, and provides high transparency and credibility for the patient management, resource scheduling and audit process of the hospital. By combining the Internet of Things technology with blockchain, the scheduling of inpatient resources has a high degree of real-time and flexibility. The resources of the hospital (such as beds, medical equipment, medical staff, etc.) can be monitored in real time through Internet of Things devices to ensure that the system can respond to changes in resource usage in a timely manner. When the resource demand or occupancy status changes, the system can automatically adjust the resource scheduling plan to achieve dynamic allocation and real-time optimization. By calculating the difference between supply and demand, the system can analyze the gap between the current supply and demand of resources in real time, and determine which resources are in short supply and which resources are in excess. At the same time, through the cost minimization constraint, the system automatically selects the resource allocation plan with the lowest cost, which not only reduces resource waste, but also reduces the overall operating cost of the hospital. After multiple iterations of optimization, the model can continuously improve the resource allocation plan to ensure that the resource scheduling plan is getting closer and closer to the optimal solution over time. By integrating patient management data, resource allocation data, real-time monitoring data, etc. into the blockchain platform, the system forms a complete closed-loop management model, which not only improves the resource management efficiency of the hospital, but also ensures that patients can receive timely care and resource support during hospitalization. The resource allocation system can respond quickly to patient needs, reduce patient waiting time, and improve patients' overall hospitalization experience and satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Other features, objects and advantages of the present application will become more apparent by reading the detailed description of non-limiting implementations made with reference to the following drawings:

[0073] Figure 1 A flowchart of a method for managing inpatients based on blockchain according to an embodiment is shown;

[0074] Figure 2 A flowchart showing the steps of a method for updating hospitalization status according to an embodiment is shown;

[0075] Figure 3 A flowchart showing the steps of a method for optimizing allocation of inpatient resources according to an embodiment is shown;

[0076] Figure 4 A flowchart of a method for managing inpatients based on blockchain according to an embodiment is shown;

[0077] Figure 5 A flowchart of the steps of a method for monitoring inpatient resource utilization according to an embodiment is shown. DETAILED DESCRIPTION

[0078] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0079] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0080] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0081] The hospital has 2,000 beds and 200 medical devices. Faced with a large number of patients admitted daily, the hospital needs an intelligent system to dynamically optimize resource allocation and ensure the reasonable scheduling of beds, equipment and medical staff. The system manages inpatients and dispatches hospital resources to improve operational efficiency and ensure efficient use of resources.

[0082] The hospital received 500 inpatient applications within a week. The system automatically obtained patient registration information through blockchain and entered the patient's basic information, hospitalization department, expected hospitalization time, etc. into the system. Patient 1: admitted to hospital, department A, expected hospitalization for 7 days, patient 2: admitted to hospital, department B, expected hospitalization for 10 days... (a total of 500 patients). The system automatically processes the registration information of these patients through blockchain technology, generates a unique identifier (UPI) for each patient, and generates initial hospitalization data. For example, the UPI of patient 1 is UPI12345, 20 beds are allocated to department A, and 30 beds are allocated to department B. The hospitalization registration and bed allocation of all patients are recorded on the blockchain, generating an allocation plan for 2,000 beds, and the hospitalization data is updated in real time through the system.

[0083] The hospital has 1,500 occupied beds and the equipment utilization rate is 75%. The 500 new patients in the hospital require an additional 500 beds. The IoT device monitors the bed utilization in real time. The bed utilization rate of Department A is 80%, and the equipment utilization rate of Department B is 85%. Bed occupancy: 1,500 beds are occupied. New demand: 500 patients need bed allocation. Medical equipment utilization: 75% equipment utilization rate. The system obtains resource occupancy data from the IoT device in real time, including the occupancy rate of current beds and equipment. The system combines the inpatient update data with the resource occupancy data and calculates the resource usage load of each department. Calculation formula: Bed occupancy update = (existing occupied beds + new patient bed demand) / total beds, bed occupancy update = (1500 + 500) / 2000 = 100% (all beds are full). The system generates data on inpatient resource occupancy, indicating that the hospital's bed resources are at full capacity and the utilization rate of medical equipment in some departments has reached a critical level, requiring optimization of resource allocation and scheduling.

[0084] The system predicts resource demand for the next week based on historical hospitalization data. In the next 7 days, 100 people will be discharged every day, and 150 new patients are expected to be admitted every day. The system also predicts that the utilization rate of medical equipment will increase to 90% in the next few days. 100 people are discharged every day, and 150 new patients are admitted every day. The utilization rate of equipment will rise to 90%. The system calculates the supply and demand difference based on the prediction algorithm and resource occupancy data, and finds that the bed will continue to be tight in the future and the equipment utilization rate is close to saturation. The system calculates the gap in future beds based on the supply and demand difference model, and prioritizes resource allocation for critically ill patients and high-demand departments through cost minimization constraints. For example, critically ill patients will be given priority to ICU beds, and equipment usage time will also be optimized to patients who need to be monitored. The system evaluates the resource utilization efficiency under different allocation strategies through simulated allocation. For example, 150 beds are allocated to critically ill patients first, and the algorithm is used to adjust the order of equipment use to improve equipment utilization. The system generates a resource optimization allocation plan, prioritizes the allocation of beds and equipment to critically ill patients, and delays hospitalization arrangements for non-emergency patients. Equipment utilization rate has been optimized to 85%, leaving a certain buffer space for the hospital.

[0085] The system generates a detailed resource scheduling plan based on the optimized allocation data, allocating 500 beds and several medical devices. At the same time, the real-time usage of all resources is monitored through IoT devices, and the scheduling process of all resources is recorded on the blockchain. Bed allocation and medical equipment usage scheduling for 500 patients. Real-time monitoring of bed usage data and equipment usage data. IoT devices collect the usage of beds and equipment in each department in real time. For example, the system detects that the bed utilization rate of Department A is 95% and the equipment utilization rate is 90%, and updates the inpatient resource scheduling plan through real-time monitoring data. If the beds in a department are idle, the system will automatically adjust the patient's scheduling order and assign new patients to the idle beds. Each resource scheduling adjustment is recorded through the blockchain to ensure that the scheduling is transparent, safe and traceable. The system generates inpatient resource scheduling plan data and adjusts resource allocation through real-time monitoring to ensure that the hospital's beds and equipment are fully utilized. The chain records the process of all resource scheduling, ensuring the transparency and security of scheduling data. Total beds: 2,000, occupied beds: 1,500, and 500 new patient needs. Medical equipment: 200 units, equipment utilization rate: 75%. Difference in bed supply and demand: The beds are full, and the demand for beds is expected to remain tight in the future. Equipment utilization rate: It is expected to reach 90% in the next week. 150 beds are allocated to critically ill patients first, and the order of equipment use is adjusted and optimized, and the equipment utilization rate drops to 85%. The Internet of Things monitors the use of beds and equipment in real time, and the system dynamically adjusts bed allocation and equipment scheduling based on real-time data. The scheduling process is recorded on the chain to ensure data transparency and non-tamperability.

[0086] See also Figures 1 to 5, the present application provides a method for inpatient management based on blockchain, comprising the following steps:

[0087] Step S1: Acquire inpatient registration data, and update the hospitalization status according to the inpatient registration data to obtain inpatient update data;

[0088] In one embodiment, the system collects basic patient information from the hospital information system (HIS) or patient registration system, including name, gender, age, reason for hospitalization, etc. At the same time, it collects additional data related to hospitalization, such as admission time, expected length of stay, department and bed allocation, etc. These data will be synchronized to the system in real time, updated as basic information, and generate the latest status of hospitalized patients. This information is recorded in a node in the blockchain network to ensure the immutability and transparency of the data.

[0089] Step S2: Processing the inpatient resource occupancy status according to the inpatient update data to obtain inpatient resource occupancy status data;

[0090] In one embodiment, the system extracts resource occupancy information such as current bed usage, medical equipment, and doctor and nurse arrangements from the patient's updated data. For each patient, the bed occupancy time, equipment usage frequency, and drug requirements are analyzed to calculate the current resource occupancy rate. By comparing the resource occupancy with the hospital's existing resources (such as available beds, number of equipment, etc.), the system derives the current resource occupancy status. The occupancy status is stored in the distributed ledger of the blockchain network to ensure transparency and provide reliable data basis for subsequent operations.

[0091] Step S3: dynamically predict resource demand and optimize allocation based on hospital resource occupancy data to obtain hospital resource optimization allocation data;

[0092] In one embodiment, the system predicts future bed, equipment and medical resource requirements based on the patient's current length of stay, disease trend and known treatment plan. Dynamically analyze other patients' discharge expectations, admission needs and transfer possibilities to predict future resource needs. Based on the predicted data, the system weighs resource allocation through internal calculations, balances supply and demand, and optimizes the allocation of existing resources to maximize resource utilization. The optimized allocation plan generates resource optimization allocation data after the algorithm simulates the rationality of scheduling.

[0093] Step S4: Implement resource scheduling and monitor resource utilization in real time according to the inpatient resource optimization allocation data to obtain inpatient resource utilization monitoring data to add to the preset inpatient management blockchain.

[0094] In one embodiment, the data of optimized allocation forms a specific scheduling plan in the system, such as which patients can transfer beds and which equipment should be reallocated to new patients. The system dispatches resources in real time according to the predetermined plan and adjusts the hospitalization arrangements and equipment use of each patient. At the same time, the system monitors the utilization of resources in real time through sensors, IoT devices and personnel feedback to ensure that resources are used efficiently. All data on resource usage and scheduling processes will be recorded in real time on the blockchain network to form an unalterable record of resource utilization, which can be reviewed and traced at any time to improve the transparency and efficiency of management.

[0095] Specifically, the system obtains patient registration data from the hospital information system (HIS), including basic information such as the patient's name, age, gender, and reason for hospitalization. Collect additional data related to hospitalization, such as: admission date: October 15, 2023, expected length of stay: 5 days, department: internal medicine, bed number: A305. The system processes these data, removes redundant or duplicate information, and updates the patient's hospitalization status. This information is recorded on the blockchain to ensure that the data cannot be tampered with and is transparent. Patient A registration data: Name: A, Age: 45 years old, Gender: Male, Reason for admission: Acute gastritis, Admission date: October 15, 2023, Estimated length of stay: 5 days, Bed: A305.

[0096] According to the registration data of patient A, the system extracts the current occupancy of resources such as beds, equipment, and medical staff. Bed A305 is occupied, and the system records its occupancy time from October 15, 2023 to October 20, 2023. The system also calculates the medical resources required daily (such as the doctor's rounds time, the number of nursing times, etc.) according to the arrangements of doctors and nurses. The system summarizes the occupancy of all resources, including beds, medical resources, and equipment usage, and generates inpatient resource occupancy data. During patient A's hospitalization (October 15 to October 20): Bed A305: occupied for 5 days Doctor: 1 round per day (5 times in total), nurse: 2 times a day (10 times in total), gastroscopy equipment used: 1 time, the system generates inpatient resource occupancy data: bed occupied for 5 days, doctor rounds 5 times, nurse care 10 times, equipment used 1 time.

[0097] The system predicts future resource needs based on the current occupancy of hospital resources. For example, the system analyzes the discharge time of other patients and the hospitalization of patient A to predict whether there will be a shortage of beds and medical resources in the future. For the hospitalization of patient A, the system considers that Zhang San is expected to stay in the hospital for 5 days and dynamically adjusts the prediction according to the development of the disease. If the condition improves, he will be discharged early; if the condition worsens, the hospitalization time will need to be extended. The system also analyzes the discharge plans of other patients in combination with the overall resource pool of the hospital to optimize resource allocation. Through the optimization algorithm, the system allocates the beds and equipment that are about to be vacated to the newly admitted patients. The hospital has another patient B who is expected to be discharged on October 18, vacating bed A306. Patient C needs to be hospitalized on October 17, so the system dynamically optimizes the bed arrangement and allocates bed A306 to patient C. If patient A's condition improves during hospitalization and he is discharged early on October 19, the system can also allocate bed A305 in advance to the next patient who needs to be hospitalized.

[0098] The system automatically schedules resources based on the optimized allocation data. For example, bed A306 is assigned to patient C, or bed A305 is assigned to a new patient after patient A is discharged early. The real-time monitoring system continuously monitors the resource utilization of the hospital through sensors and IoT devices to ensure that beds, medical resources, equipment, etc. are used reasonably as planned. The system records all scheduling and resource utilization data and uploads the data to the blockchain to ensure that the information cannot be tampered with, which is convenient for subsequent auditing and tracing. Patient A was discharged early on October 19, 2023, and bed A305 was vacant. The system reallocated bed A305 to the newly admitted patient D. Both scheduling information and real-time monitoring data are uploaded to the blockchain to ensure information transparency and easy traceability.

[0099] Preferably, step S1 specifically includes:

[0100] Step S11: Obtain patient identity data, and perform electronic file query based on the patient identity data to obtain the patient's electronic file data;

[0101] In one embodiment, the system obtains data from the identity information provided by the patient, including ID number, name, gender, age, etc. The system is connected to the hospital electronic archive system or the national medical archive database, and enters the patient's identity information to query the patient's historical electronic medical records or related health data. The system automatically retrieves the patient's electronic archive data, including past medical history, allergy history, medication records, etc. Through the patient's ID number or social security number, the system retrieves the patient's electronic file from the database. The system matches and verifies the patient's health data to ensure that the identity information is correct. The search results are fed back in the form of a data structure, including the patient's basic information, past medical history, etc.

[0102] Step S12: Generate a unique patient identification according to the patient's electronic file data to obtain the patient's unique identification data;

[0103] In one embodiment, the system extracts key information from the electronic archive data, such as fields such as ID number, date of birth, and gender, as input for generating a unique identifier. The extracted information is encrypted and hashed using a hash function to generate a unique identifier for the patient. The unique identifier serves as the patient's identity on the blockchain for subsequent registration and information tracing. The system extracts several fields from the patient's electronic archive, such as ID number, name, etc. The system combines and inputs these fields into the encryption algorithm to generate a unique hash value. The generated unique identifier is associated with other patient information and stored as the basis for subsequent processing. Patient A's admission information: admission date: October 18, 2023, department: Internal Medicine, ward: 302, bed: A305, on-chain registration data: unique identifier: HASH_ID_1A2B3C4D5E6F7G8H9I, hospitalization registration information: admission date (October 18, 2023), department (Internal Medicine), bed (A305).

[0104] Step S13: Perform on-chain registration records based on the patient's unique identification data to obtain inpatient registration data;

[0105] In one embodiment, the system registers the patient's unique identification data into the distributed ledger of the blockchain to form a unique identity record on the blockchain. During registration, the system associates the patient's unique identifier with the hospitalization information such as the time of admission, department, and bed and stores it on the chain. Once the registration record is written to the blockchain, it cannot be tampered with, ensuring that the patient's hospitalization information is consistent with the identity information. The system calls the interface of the blockchain node, packages the patient's unique identifier and hospitalization information together, and generates a transaction record. Each node in the blockchain network verifies the legitimacy of the record, ensures that the data is accurate, and then reaches consensus. After consensus, the record is successfully written into the blockchain and becomes an unalterable data block.

[0106] Step S14: generating initial hospitalization data according to the inpatient registration data to obtain initial hospitalization data;

[0107] In one embodiment, the system automatically generates initial hospitalization information based on the inpatient data registered on the chain, including the patient's bed allocation, department information, admission time, etc. The system will also automatically retrieve relevant treatment plans or historical cases (fuzzy search for similar situations) based on the patient's electronic file data to prepare for the estimation of resource calls for the current hospitalization situation. The initial data is updated synchronously in the internal system and the blockchain as the basic record of the patient's hospitalization.

[0108] Step S15: Perform status update confirmation based on the initial hospitalization status data to obtain updated inpatient data.

[0109] In one embodiment, the system compares and verifies the initial hospitalization data with the patient's actual admission situation to confirm whether the bed allocation and medical care arrangements are correct. Based on the confirmed admission situation, the system updates the patient's hospitalization status data, such as whether the patient has been admitted to the bed, whether treatment has begun, etc. The updated inpatient status data is registered again through the blockchain to ensure that each status update can be accurately traced. The system checks the patient's actual hospitalization situation, including whether the admission is as planned and whether it matches the bed and doctor. If there are any adjustments or changes, such as changes in wards or changes in treatment plans, the system will be updated synchronously. After each status update, the system generates a new block record and uploads the updated data to the blockchain to ensure that the status change is traceable.

[0110] Preferably, step S2 specifically includes:

[0111] Step S21: Obtaining hospitalization resource occupancy data;

[0112] In one embodiment, the system extracts the resource usage related to the current inpatient from the hospital's resource management system, including: bed resources: the beds occupied by the patient, the duration of bed use, the number of available beds, etc. Medical resources: the frequency of patient care, the doctor's rounds, the arrangement of doctors and nurses, etc. Equipment resources: the usage of medical equipment used by patients (such as ECG monitors, gastroscopes, CT machines, etc.), the duration of equipment occupancy, etc. The system automatically retrieves the real-time usage of these resources based on the information of the inpatient (such as bed number, department, etc.). The system extracts relevant information such as beds, medical staff, and medical equipment from the resource management system to form inpatient resource occupancy data. The data includes the bed number of each patient, the duration of medical care used per day, the number of used medical equipment, etc. Resources occupied by patient A: Bed: Bed A305, occupied for 4 days. Medical resources: doctors make rounds once a day, and nurses provide care twice a day. Equipment resources: gastroscope equipment is used once, and ECG monitor is used for 4 consecutive days.

[0113] Step S22: Analyze the inpatient resource occupancy situation according to the inpatient resource occupancy data and the inpatient update data to obtain inpatient resource occupancy analysis data;

[0114] In one embodiment, the system matches the acquired inpatient resource occupancy data with the updated data of the inpatients, and analyzes whether the current resources are in a normal utilization state, and whether there is a waste of resources or insufficient resources. The system performs analysis based on the patient bed utilization rate, the amount of medical resources invested, and the equipment utilization rate. For example: Bed utilization rate analysis: whether the bed is idle for a long time or occupied overtime. Medical resource matching analysis: whether the workload of doctors and nurses is balanced. Equipment usage analysis: whether the occupancy rate of medical equipment is reasonable, and whether there is a queue waiting. The system matches the updated data of inpatients (such as hospitalization time, condition, etc.) with the resource utilization to confirm whether the resource utilization is consistent with the actual needs of the patients. The system uses a data calculation model to compare the supply and demand of resources and generate resource occupancy analysis data. For example, indicators such as the patient bed vacancy rate, the frequency of use of medical resources, and the length of time the equipment is used will all be involved in the analysis. Analysis of inpatient resource occupancy of patient A: Bed A305: occupied for 4 days, normal use. Medical resources: the frequency of ward rounds and nursing meets the requirements, and medical resources are reasonably matched. Equipment resources: The gastroscope equipment is used once, and the ECG monitor is used continuously for 4 days, with no waste of equipment.

[0115] Step S23: Record the resources on the chain according to the inpatient resource occupancy analysis data to obtain inpatient resource occupancy data.

[0116] In one embodiment, the system registers the resource occupancy analysis data into the blockchain to ensure the transparency and immutability of the data and provide a reliable basis for subsequent resource optimization and scheduling. The system records the detailed usage of resources, such as the bed number occupied by the patient, the usage time of the equipment, the working hours of medical staff, and other data, and uploads them to the distributed ledger of the blockchain network. The system packages the resource occupancy analysis results, generates a transaction record on the chain, and confirms it through the consensus mechanism of the blockchain network. Once the data is recorded on the chain, a permanent resource usage record is formed, and any subsequent resource scheduling or resource allocation adjustment can be based on this record. Patient A's hospitalization resource occupancy chain record: Bed A305: occupied for 4 days. Medical resources: Doctors make rounds once a day, nurses provide care twice, and all are reasonably allocated. Equipment resources: gastroscopy equipment is used once, and ECG monitors are used for 4 consecutive days. These data are stored in the blockchain and can be traced and audited in the future.

[0117] Preferably, step S3 is specifically:

[0118] Step S31: extracting characteristics of the inpatient resource occupancy situation according to the inpatient resource occupancy situation data to obtain inpatient resource occupancy situation characteristic data;

[0119] In one embodiment, the system extracts key features of resource usage from existing inpatient resource occupancy data. Common features include: bed occupancy rate: bed occupancy time and idle time. Medical staff load: the time and workload allocated by doctors and nurses to different patients. Equipment utilization rate: the usage time, frequency and queue of medical equipment (such as monitors, CT machines, etc.). The system normalizes the data and converts the original data into feature data suitable for analysis and prediction to ensure that the feature data can describe the overall occupancy of resources. The system traverses the inpatient resource occupancy data and extracts the key feature values ​​of beds, medical resources and equipment through screening and calculation. The data processing process includes the calculation of idle rate, utilization rate and frequency to generate a feature data table. Patient A: Bed occupancy rate: 100% (occupied for 4 days, no idle time). Doctor load: 1 round per day, 4 times in total. Nurse load: 2 rounds per day, 8 times in total. Equipment utilization rate: Gastroscope: used once, utilization rate 25% (based on 4 days of hospitalization). ECG monitor: Continuous use for 4 days, utilization rate 100%.

[0120] Step S32: forecasting the demand for inpatient resources based on the characteristic data of inpatient resource occupancy to obtain forecast data for inpatient resource demand;

[0121] In one embodiment, the system predicts the demand for future resources based on the extracted resource occupancy feature data by means of historical trend analysis, time series analysis, etc. The system will refer to multiple factors such as the patient's historical hospitalization time, the severity of the disease, the resource utilization efficiency of the current hospitalization, etc., to generate forecast data for future resource demand. The forecast content includes the vacancies and demand for future beds, the workload of doctors and nurses, and the use of equipment. Based on the patient's hospitalization days, medical needs and equipment use, the system predicts whether the patient will continue to occupy resources in the next few days, or whether there will be resource needs for new patients. Data calculation includes the changing trend of bed demand rate and equipment utilization rate. Demand forecast for patient A: Bed: It is expected to stay in hospital for another day, and the bed utilization rate will remain 100%. Medical staff: In the next day, the doctor will make rounds once and the nurse will provide care twice. Equipment demand: There will be no new equipment demand in the next day, and the ECG monitor will continue to be used for 1 day.

[0122] Step S33: constructing an inpatient resource optimization allocation model according to the inpatient resource demand forecast data to obtain an inpatient resource optimization resource allocation model;

[0123] In one embodiment, the system constructs an inpatient resource optimization allocation model based on demand forecast data, and the optimization goals include: maximizing resource utilization: reducing idle resources by reasonably arranging the use of beds and equipment. Balancing medical care load: evenly distributing the workload of doctors and nurses to avoid excessive concentration on certain patients or departments. Minimizing equipment waiting time: reducing the queuing or idle time of equipment by adjusting the order or arrangement of equipment use. The model ensures that hospital resources can be effectively dispatched to meet the needs of different patients through resource supply and demand balance calculation.

[0124] The system uses the forecast data as input and sets optimization goals and constraints, such as maximizing bed utilization and balancing doctors' workload. The system executes the optimization algorithm to generate the optimal allocation plan. The optimization model generates the goal of maximizing bed utilization, equipment utilization efficiency, and balancing medical resources. Input patient A's bed demand, medical care demand, and equipment demand for the next day. Output the adjusted bed, medical care, and equipment allocation plan.

[0125] Step S34: generating an inpatient resource optimization plan according to the inpatient resource optimization resource allocation model to obtain inpatient resource optimization plan data;

[0126] In one embodiment, the system generates a specific resource allocation plan based on the optimization allocation model, including: bed arrangement, how to allocate vacant beds to new patients, or whether the current patient needs to transfer beds. Adjustment of medical staff, whether the shifts of doctors and nurses need to be adjusted, and whether the nursing frequency of certain patients needs to be increased or decreased. Equipment scheduling, whether the use priority and use time of equipment need to be rearranged to avoid equipment idle or waiting. The system generates a specific allocation plan based on the allocation results in the optimization model, including bed allocation, medical staff scheduling and equipment use plan. The optimization plan is implemented and monitored in the hospital management system. Optimization plan for patient A: Bed allocation: Bed A305 continues to be allocated. Doctor's rounds: continue to be maintained once a day. Nurse care: continue to be maintained twice a day. Equipment scheduling: ECG monitor continues to be used for 1 day, and there is no need to adjust the equipment arrangement.

[0127] Step S35: Perform on-chain allocation records according to the inpatient resource optimization plan data to obtain inpatient resource optimization allocation data.

[0128] In one embodiment, the system uploads the generated optimization plan to the blockchain, records the optimization allocation process, and ensures the transparency and traceability of the optimization allocation. The system stores the specific data in the optimization allocation plan (such as patient bed allocation, medical staff adjustment, equipment use plan, etc.) in a distributed ledger to ensure the security and immutability of the data. The system packages the allocation data of the optimization plan through the blockchain interface, generates block transactions and uploads them to the chain. The blockchain node confirms the transaction through the consensus mechanism and records the optimization allocation data of inpatient resources. Patient A's on-chain allocation record: Bed: A305, the allocation time is extended by another day. Medical staff: Doctors make ward rounds once and nurses provide care twice. Equipment: ECG monitor is used continuously for 1 day. These data are stored in the blockchain and can be audited at any time.

[0129] Preferably, the feature extraction of hospitalization resource occupancy is specifically as follows:

[0130] Perform data repair according to the inpatient resource occupancy data to obtain inpatient resource occupancy repair data;

[0131] In one embodiment, the system checks whether there are missing values, outliers or data errors in the inpatient resource occupancy data. Data repair methods include: filling missing values, using the mean, median or interpolation method to fill in the missing data. Deleting outliers, identifying and removing data points that exceed a reasonable range, such as occupancy time or medical resources that exceed the normal range. Data correction, correcting erroneous records according to business rules, such as conflicts where the same device is used by multiple patients at the same time. The system first performs an integrity check on the original data and marks missing data and abnormal data. The system then uses a preset strategy to fill or correct this data to generate clean, usable inpatient resource occupancy repair data. The occupancy time record of bed A305 is missing, and it is filled with the average occupancy time of adjacent days. The usage time of a certain CT device is abnormally high, and the system adjusts it according to normal working hours.

[0132] Perform target coding according to the inpatient resource occupancy repair data to obtain inpatient resource occupancy target coding data;

[0133] In one embodiment, the system encodes categorical data (such as bed number, department, and equipment type) for subsequent numerical calculations. Target encoding methods include: Ordinal encoding, mapping categorical features (such as bed number) to specific numerical values. Average target encoding, calculating the average value of each classification and encoding it based on the relationship between the feature and the target variable. The system encodes the categorical information (such as bed number, equipment type) in the repaired inpatient resource data to convert these data into numerical features that can be used for algorithm processing. Bed A305 is coded as 305, and the equipment CT scanner is coded as 001.

[0134] Performing hierarchical feature extraction based on the target coding data of inpatient resource occupancy to obtain the first inpatient resource occupancy characteristic data;

[0135] In one embodiment, the system stratifies the resources by resource type (such as beds, medical care, and equipment) based on the encoded data, and extracts the key features of each layer. The extracted features include: bed layer, extracting features such as bed occupancy time, idle rate, and turnover rate. Medical care layer, extracting features such as doctors' and nurses' workload and ward round frequency. Equipment layer, extracting features such as equipment usage time, usage frequency, and waiting time. The system stratifies resources by type, and calculates the data features of each layer separately to generate the first inpatient resource occupancy feature data. For example, bed A305: occupancy time is 4 days, idle rate is 0%, and turnover rate is high. Doctor workload: ward rounds once a day, nurses provide care twice a day.

[0136] Cluster calculation is performed based on the target coding data of hospitalization resource occupancy to obtain cluster data of hospitalization resource occupancy;

[0137] In one embodiment, the system performs cluster analysis on the encoded inpatient resource occupancy data, and the clustering algorithm can classify inpatients or resources with similar resource usage patterns into the same category. Through clustering, common or special patterns of resource occupancy can be identified. Based on the frequency, duration, load and other characteristics of resource usage, the system calculates the similarity of different patients or resources, and clusters objects with high similarity into one category. The system clusters the data of each bed, equipment, and medical resource to form different categories of resource occupancy, which facilitates the identification of resource usage patterns. Inpatients are divided into three categories: high occupancy group: beds are occupied for a long time and equipment is used frequently. Medium occupancy group: beds are occupied for a moderate time and equipment is used at an average frequency. Low occupancy group: beds and equipment are used at a low frequency.

[0138] Extract cluster distribution features of the cluster data of hospitalization resource occupancy to obtain characteristic data of the second hospitalization resource occupancy situation;

[0139] In one embodiment, the system extracts the distribution characteristics of each cluster based on the clustering results, including the center point of the cluster, the distribution density of the data points in each cluster, the boundary conditions of the cluster, etc. By analyzing the distribution characteristics of each cluster, the resource usage pattern can be better understood. The center point of the cluster represents the typical pattern of resource usage in the cluster. The cluster density reflects the concentration of data points in the cluster. The system calculates the distribution characteristics of each cluster after clustering, including the center, density and boundary conditions, and generates characteristic data of the second inpatient resource occupancy. The high occupancy group has a high resource occupancy density, the beds in the center of the cluster are used for a long time, and the medical workload is heavy. The medium occupancy group has relatively average resource occupancy, and the data distribution within the cluster is relatively dispersed.

[0140] Perform characteristic density calculation according to the first inpatient resource occupancy characteristic data and the second inpatient resource occupancy characteristic data, to obtain the first inpatient resource occupancy characteristic density data and the second inpatient resource occupancy characteristic density data respectively;

[0141] In one embodiment, the system calculates the density distribution of each feature based on the first and second feature data, and the density reflects the distribution of the feature in the entire data set. A feature with high density indicates that the feature is more important or frequent in inpatient resource management. A feature with low density indicates that the feature is relatively minor or rare in inpatient resource management. The system calculates the density of the feature based on the distribution of each feature. Features with high density mean that resources are frequently used, and features with low density mean that they are used less frequently. Bed occupancy density: high density (high bed utilization, tight resources). Equipment usage density: medium density, moderate frequency of equipment usage.

[0142] Performing feature mutual information analysis based on the first inpatient resource occupancy feature data and the second inpatient resource occupancy feature data to obtain inpatient resource occupancy feature mutual information data;

[0143] In one embodiment, the system evaluates the correlation between features by calculating the mutual information between each feature. Mutual information reflects the amount of information a feature provides when understanding other features. Features with high mutual information indicate that the features are highly correlated. Features with low mutual information indicate that the features are relatively independent. The system calculates the mutual information between each pair of features to evaluate the correlation of the features. Features with high mutual information may be used for the same type of resource optimization, and features with low mutual information can be optimized independently. Mutual information between bed occupancy and equipment use: high mutual information, indicating that bed occupancy is associated with the frequency of equipment use. Mutual information between medical care load and equipment use: medium mutual information, with a certain correlation.

[0144] According to the first inpatient resource occupancy characteristic density data, the second inpatient resource occupancy characteristic density data and the inpatient resource occupancy characteristic mutual information data, nonlinear feature fusion is performed on the first inpatient resource occupancy characteristic data and the second inpatient resource occupancy characteristic data to obtain the inpatient resource occupancy characteristic data.

[0145] In one embodiment, the system nonlinearly fuses the first and second feature data to generate comprehensive resource occupancy feature data. Nonlinear fusion can capture the complex nonlinear relationship between features and improve the expressiveness of the model. The system uses feature density and mutual information to integrate the weights of different features and generate new features through nonlinear combination. The system combines feature density and mutual information, uses a nonlinear fusion algorithm (such as weight calculation based on density features and feature mutual information), and fuses all extracted feature data into a feature data set to ensure that the complex relationship of resource use can be captured in the optimization model. Features such as bed usage, medical load, and equipment usage are integrated to generate comprehensive resource occupancy feature data for optimization models.

[0146] Preferably, the characteristic density is calculated as follows:

[0147] Estimating the distribution type according to the first inpatient resource occupancy characteristic data and the second inpatient resource occupancy characteristic data, to obtain first characteristic distribution type estimation data and second characteristic distribution type estimation data respectively;

[0148] In one embodiment, the system estimates the distribution type of each feature based on the statistical distribution of the feature data. Distribution types include normal distribution, exponential distribution, uniform distribution, etc. Histogram analysis observes the distribution form of the data by drawing a histogram of the feature data, and preliminarily determines the distribution type of the feature data. Parameter fitting By fitting different distribution functions, the system can calculate the goodness of fit of each distribution (such as the least squares method) and determine the optimal distribution type. The system draws a histogram based on the characteristic data of the first and second inpatient resource occupancy conditions, calculates the goodness of fit of each distribution type, and finally estimates the best distribution type. The bed occupancy time feature is observed through a histogram, and the bed occupancy time is normally distributed. Equipment usage frequency feature The equipment usage frequency is approximately uniformly distributed.

[0149] Using the first characteristic distribution type estimation data, the first inpatient resource occupancy characteristic data is labeled with a distribution type to obtain first characteristic distribution type labeled data, and using the second characteristic distribution type estimation data, the second inpatient resource occupancy characteristic data is labeled with a distribution type to obtain second characteristic distribution type labeled data;

[0150] In one embodiment, the system labels the data of each feature according to the estimated distribution type. The labeling step is to assign each feature its corresponding distribution type for subsequent distribution fitting and kernel density estimation. The labeling process identifies the feature data according to the estimated distribution type, such as normal distribution is labeled as "Normal", uniform distribution is labeled as "Uniform", etc. The system adds a corresponding distribution type label to each feature data for subsequent analysis steps. The bed occupancy time feature is labeled as Normal distribution. The equipment usage frequency feature is labeled as Uniform distribution.

[0151] Perform kernel density estimation and distribution fitting based on the first feature distribution type labeled data and the second feature distribution type labeled data to obtain first feature distribution density data and second feature distribution density data;

[0152] In one embodiment, the system uses kernel density estimation to smooth the feature data and obtain the probability density function of the feature distribution. Kernel density estimation is used to fit feature data whose distribution is not completely symmetrical or non-standard. During the kernel density estimation process, the system automatically selects the appropriate bandwidth according to the distribution characteristics of the data to ensure the smoothness and accuracy of the estimation results. The system labels the data according to the feature distribution type, uses the kernel density estimation method to fit the distribution of the feature data, and generates a feature density function. Kernel density estimation of bed occupancy time: normal distribution fitting, bandwidth h=0.5. Kernel density estimation of equipment usage frequency: uniform distribution fitting, bandwidth h=0.8.

[0153] Density calculation is performed based on the first characteristic distribution density data and the second characteristic distribution density data to obtain first inpatient resource occupancy characteristic density data and second inpatient resource occupancy characteristic density data, respectively.

[0154] In one embodiment, the system calculates the distribution density of each feature based on the results of kernel density estimation. The density value reflects the distribution of the feature in a specific area, and a higher density indicates that the feature is more concentrated in the area. By integrating the probability density function, the system can calculate the distribution density of the feature in different intervals. Based on the kernel density estimation results of each feature, the system obtains the density value of each feature through integral calculation, and stores it as the first and second inpatient resource occupancy feature density data. The bed occupancy time density is calculated to have a density value of 0.25, indicating that the distribution concentration of the feature in a specific interval is relatively high. The equipment usage frequency density is calculated to have a density value of 0.15, indicating that the frequency of use of the feature is relatively dispersed in a certain interval.

[0155] Preferably, the nonlinear feature fusion is specifically as follows:

[0156] A density-weighted feature mutual information matrix is ​​generated according to the first inpatient resource occupancy characteristic density data, the second inpatient resource occupancy characteristic density data, and the inpatient resource occupancy characteristic mutual information data to obtain density-weighted feature mutual information matrix data;

[0157] In one embodiment, the system constructs a weighted mutual information matrix based on the density data of the first and second features, and the mutual information data between the features. Feature density data is used to weight the mutual information, and the higher the density of the feature, the greater the weight in the mutual information. The density-weighted mutual information matrix can better reflect the strength of the association between the features, while taking into account the distribution of the features in the data. Each element in the mutual information matrix represents the mutual information value between two features, reflecting their mutual dependence. The system first calculates the mutual information matrix between the features, and uses the density data of the features to weight each mutual information value. The generated density-weighted mutual information matrix is ​​used to describe the strength of the interaction between the features. Mutual information between bed occupancy time and equipment usage frequency: higher density and larger mutual information value indicate that the two features have a stronger correlation. Mutual information between medical workload and equipment usage: lower density and smaller mutual information value indicate weaker correlation.

[0158] Constructing a feature interaction network for the first inpatient resource occupancy feature data and the second inpatient resource occupancy feature data according to the density-weighted feature mutual information matrix data to obtain feature interaction network data;

[0159] In one embodiment, the system constructs a feature interaction network based on a weighted mutual information matrix. Each node represents a feature, and the weight of the edge represents the strength of interaction between features (i.e., the weighted mutual information value). The connection and weight between the nodes of the network graph model reflect the interactive relationship between the features, and the system constructs a nonlinear association between each feature and other features through a network graph. The system generates a weighted network graph model based on the data in the mutual information matrix. The relationship between the features is described as network nodes and edges, and the weight of the edge is determined by the mutual information value. Bed occupancy time and equipment use frequency: connected by high-weight edges in the network, indicating a strong interactive relationship. Medical workload and equipment use: The connection is weak and the interaction is not significant.

[0160] A regression model and a neural network are constructed according to the feature interaction network data to obtain a first nonlinear interaction feature model and a second nonlinear interaction feature model respectively;

[0161] In one embodiment, the system first uses the relationship in the feature interaction network to build a regression model to capture the linear and nonlinear relationships between features. The regression model generates a series of interactive features based on the structure of the feature interaction network. The system automatically selects feature combinations with large mutual information, builds a linear regression model, and calculates the linear relationship between these features. In order to capture complex nonlinear relationships, the system further builds a neural network model based on the feature interaction network. Neural networks can capture more complex nonlinear feature interactions through multiple hidden layers. The system uses data from the feature interaction network to train the neural network model to learn the complex nonlinear relationships between features. The system first builds a regression model based on the feature interaction network to extract linear and partially nonlinear relationships. Subsequently, a deep nonlinear model is further built through the neural network to learn more complex feature interactions. The regression model found that there is a significant linear relationship between bed occupancy time and equipment usage frequency. The neural network model found a nonlinear association between medical workload and equipment usage frequency.

[0162] Nonlinear interaction features are generated according to the first nonlinear interaction feature model and the second nonlinear interaction feature model to obtain first nonlinear interaction feature data and second nonlinear interaction feature data respectively.

[0163] In one embodiment, the system generates new nonlinear interaction features based on the parameters in the regression model and the neural network model. These features reflect the complex interactions in the resource occupation process that are not easily captured by the linear model. The first nonlinear interaction feature comes from the interaction feature of the regression model, which mainly captures linear relationships and some weak nonlinear relationships. The second nonlinear interaction feature comes from the complex nonlinear interaction feature of the neural network model, which can capture multi-level and multi-dimensional nonlinear relationships. The system generates a new feature set by calculating the outputs of the first and second interaction feature models to describe the various nonlinear interactions in resource occupation. The first nonlinear interaction feature is generated by the regression model, representing the linear and partially nonlinear interactions between the bed occupancy time and the frequency of equipment use. The second nonlinear interaction feature is generated by the neural network, representing complex multi-dimensional nonlinear interactions, such as the deep relationship between medical care load and equipment use.

[0164] Performing nonlinear feature fusion on the first inpatient resource occupancy condition feature data and the second inpatient resource occupancy condition feature data according to the first nonlinear interactive feature data to obtain first occupancy condition nonlinear feature data;

[0165] In one embodiment, weighted sum is used to perform weighted fusion of features. The advantage of the weighted sum method is that it can dynamically adjust the weight of each feature according to its importance, such as calculating the mutual information value between the first nonlinear interaction feature data and the first inpatient resource occupancy feature data, or evaluating the correlation between the first nonlinear interaction feature data and the first inpatient resource occupancy feature data based on the coefficient of the regression model. If the first nonlinear interaction feature data and the first inpatient resource occupancy feature data have a higher explanatory power, then the first inpatient resource occupancy feature data can be given a higher weight in the weighted sum, and the second inpatient resource occupancy feature data is processed in the same way. The feature weight is adjusted by the first nonlinear interaction feature data, so that the first inpatient resource occupancy feature data and the second inpatient resource occupancy feature data are weighted and calculated to obtain the first occupancy nonlinear feature data.

[0166] According to the second nonlinear interactive feature data, nonlinear feature fusion is performed on the first inpatient resource occupancy feature data and the second inpatient resource occupancy feature data to obtain second occupancy nonlinear feature data;

[0167] In one embodiment, an activation function is used to perform nonlinear mapping on the features, and a specific weight value of the activation function is determined by the second nonlinear interaction feature data. The second nonlinear interaction feature data, the first inpatient resource occupancy feature data, and the second inpatient resource occupancy feature data are used as inputs and passed into a nonlinear activation function (such as ReLU, Sigmoid, etc.). The fusion process of the second nonlinear interaction feature data, the first inpatient resource occupancy feature data, and the second inpatient resource occupancy feature data is nonlinearly mapped by the activation function, so that the second occupancy nonlinear feature data with nonlinear features can be generated.

[0168] The first occupancy situation nonlinear characteristic data and the second occupancy situation nonlinear characteristic data are arranged in layers to obtain the inpatient resource occupancy situation characteristic data.

[0169] In one embodiment, the first occupancy situation nonlinear feature data and the second occupancy situation nonlinear feature data are hierarchically arranged. Hierarchical arrangement refers to sorting and merging two feature sets, and arranging them hierarchically according to specific rules or logic to ensure that the result can maximize the retention of the hierarchy of the data and the relationship between the features. For the first occupancy situation nonlinear feature data and the second occupancy situation nonlinear feature data, the data are merged and arranged by a sorting algorithm (such as sorting based on importance or sorting based on feature relevance) to obtain the inpatient resource occupancy situation feature data.

[0170] Preferably, the construction of the inpatient resource optimization allocation model is specifically as follows:

[0171] Calculate the difference between supply and demand of inpatient resources based on the forecast data of inpatient resource demand, and obtain the difference between supply and demand of inpatient resources;

[0172] In one embodiment, the system calculates the supply and demand difference of each resource based on the predicted inpatient resource demand data and the current resource supply data. The supply and demand difference indicates the gap or surplus between the hospital's existing resources and the predicted demand. Supply and demand difference = predicted resource demand - current resource supply. If the supply and demand difference is positive, it means that there is a shortage of resources; if it is negative, it means that there is an excess of resources. The supply and demand difference can be applied to various resources, including beds, equipment, medical staff, etc. The system extracts the resource data required for each department or patient from the prediction model, and compares it with the current available resources of the hospital to generate supply and demand difference data. Bed supply and demand difference, ward A is expected to require 20 beds, but currently only 15 beds are available, and the supply and demand difference is 5 (indicating a shortage). Equipment use supply and demand difference, CT equipment is required 3 times, but the existing equipment can meet 2 times of demand, and the supply and demand difference is 1 (indicating a shortage).

[0173] According to the difference data of hospitalization resource supply and demand, the hospitalization resource cost minimization constraint is performed to obtain the hospitalization resource supply optimization data;

[0174] In one embodiment, the system performs cost minimization calculations based on the supply and demand differences of inpatient resources to ensure that the cost of resource allocation is reduced while meeting resource needs. Costs include direct use costs (such as equipment use costs, personnel costs) and indirect costs (such as resource waste, extended patient waiting time, etc.). Each type of resource has its corresponding cost function, and the system will minimize the cost of resource use through optimization algorithms (such as linear programming) while meeting the supply and demand balance. The system optimizes the supply and demand difference data based on the cost model of resource use, and calculates the optimal resource supply plan to minimize the cost of resource allocation. Minimize the cost of bed allocation. The bed use costs of different wards are different. The system calculates to find beds that can effectively fill the supply and demand gap and have the lowest total cost. Minimize the cost of equipment use. The system allocates the optimal equipment use time by balancing the frequency of equipment use and related costs, reducing unnecessary equipment idle or waiting time.

[0175] Prioritize hospitalization resources according to the hospitalization resource supply optimization data to obtain hospitalization resource priority ranking data;

[0176] In one embodiment, the system determines the allocation priority of each type of resource based on the data after resource supply and demand optimization. Priority sorting takes into account the following factors: urgency of demand, giving priority to meeting the needs of urgent conditions or resource shortages. Resource utilization, giving priority to resources with higher utilization rates to reduce waste. Cost-effectiveness, giving priority to low-cost, high-efficiency resources. The system uses a multi-standard decision model and combines multiple dimensions to prioritize resources. The system analyzes the urgency, utilization efficiency and other data of different resources, generates resource priority data through a weighted sorting method, and determines which resources should be allocated first. Bed priority sorting, the bed demand of patients with urgent needs has the highest priority, and beds are allocated first. Equipment use priority sorting, equipment is allocated to patients in urgent need first to avoid long waiting times.

[0177] In one embodiment, the system optimizes resource allocation based on the supply and demand difference data to ensure that the resource demand is met while reducing the cost of resource use and allocation. Costs include: direct costs such as bed allocation costs, equipment use costs, and medical staff scheduling costs. Indirect costs include patient waiting time, inefficient use costs caused by idle resources, etc. The system uses optimization algorithms, such as linear programming or dynamic programming, to minimize the total resource cost, ensuring that the optimized allocation results can meet the demand as much as possible and reduce the cost of resource allocation. The system establishes a cost function, inputs the supply and demand difference data, and uses the optimization algorithm to find the lowest cost allocation plan. For example, the system may choose to reallocate some wards with higher idle rates, or reduce personnel costs by adjusting the doctor and nurse scheduling. Minimize bed cost: The cost of a single bed in ward A is $50 / day, and the cost of a bed in ward B is $40 / day. The system chooses to prioritize the allocation of beds in ward B (lower cost) to meet demand, while keeping the beds in ward A as a backup. The system ensures that the equipment operates efficiently by balancing the frequency of equipment use with the associated maintenance costs, avoiding high costs caused by under- or over-use.

[0178] Based on the priority ranking data of inpatient resources, simulated allocation and iterative optimization are carried out to build an inpatient resource optimization resource allocation model.

[0179] In one embodiment, the system simulates allocation based on the priority ranking data of resources. During the simulated allocation process, different resources are gradually allocated to patients or departments in order of priority. After allocation, the system will rebalance supply and demand according to the allocation results. If it is found that some resource allocations are unreasonable or new supply and demand differences appear, the system will readjust and iteratively optimize. The system uses iterative algorithms (such as greedy algorithms, genetic algorithms, etc.) for step-by-step optimization, and continuously adjusts the resource allocation plan until the supply and demand difference is minimized and the resource utilization efficiency is maximized. The system simulates the allocation of each type of resource, observes the supply and demand status after resource allocation, and adjusts it according to the results, iteratively optimizes the allocation plan until the supply and demand of resources are balanced and the cost is minimized. Bed simulation allocation, in which beds are first allocated to the most urgent patients, and the system then adjusts the bed allocation of other patients until the needs of all patients are met. Equipment simulation allocation, the use of CT equipment is allocated to emergency patients according to demand priority, and the utilization rate of equipment is further optimized through iteration to ensure that all needs are met.

[0180] Preferably, step S4 is specifically:

[0181] Step S41: generating an inpatient resource scheduling plan according to the inpatient resource optimization allocation data to obtain inpatient resource scheduling plan data;

[0182] In one embodiment, the system formulates a detailed resource scheduling plan based on the inpatient resource optimization allocation data, and the specific contents include: bed scheduling, which beds should be allocated to which patients, the scheduling time window, the expected occupancy time, etc. Equipment scheduling, such as the use time arrangement of CT equipment, monitors, etc., to avoid conflicts or long waiting times. Personnel scheduling, including personnel scheduling arrangements such as doctor rounds and nurse care, ensures the rational use of medical resources. The scheduling plan takes into account multiple factors such as the urgency, availability, and cost-effectiveness of resources to ensure the efficient use of resources in different time periods. The system extracts the allocation data of beds, equipment, and medical staff from the optimization allocation results, and generates a resource scheduling plan arranged in time sequence to ensure that the right resources are provided at the right time and the right place. The bed scheduling plan is specifically that the bed in ward A is allocated to patient Zhang San from 10:00 to 14:00 on October 20, and the estimated occupancy time is 4 hours. The equipment scheduling plan is specifically that the CT equipment will be allocated to patient Li Si at 14:30 on October 20, and the estimated usage time is 1 hour. The specific personnel scheduling plan is that the nurse will provide care for Zhang San at 10:30 on October 20, and the doctor will conduct ward rounds at 11:00.

[0183] Step S42: collecting and adjusting IoT data according to the inpatient resource scheduling plan data to obtain inpatient resource scheduling collection data;

[0184] In one embodiment, the system monitors the use of inpatient resources in real time through the Internet of Things (IoT) devices within the hospital, such as the occupancy status of beds, the operation of equipment, the presence of medical staff, etc. IoT sensors and monitoring devices collect the occupancy status of wards and equipment in real time, and automatically record resource usage and status changes (such as the start or end of equipment use, patients entering and leaving beds, etc.). The system adjusts the scheduling plan in real time based on the actual data collected by the IoT. If conflicts or idle resources are found in resource scheduling, the system will automatically make adjustments. For example, if the use time of a certain device ends early, the system will arrange for the next patient to use the device early. The system monitors the actual status of each resource through IoT devices and compares it with the scheduling plan. If the actual situation does not match the plan, the system will dynamically adjust the scheduling plan and update the scheduling plan in real time.

[0185] Step S43: Collect data based on the inpatient resource scheduling and perform on-chain monitoring and recording to obtain inpatient resource utilization monitoring data.

[0186] In one embodiment, the system records the real-time use of inpatient resources through blockchain to ensure the transparency and immutability of resource use. Monitoring records include: bed usage monitoring, recording the actual occupancy time, idle time, patient entry and exit records of each bed, etc. Equipment usage monitoring, recording the start time, end time, usage frequency and maintenance of the equipment. Personnel work monitoring, monitoring the actual execution of activities such as doctor rounds and nurse care, to ensure that medical staff perform tasks as planned. The system uploads these monitoring data to the blockchain to ensure the integrity and traceability of the data.

[0187] Preferably, the present application also provides a blockchain-based inpatient management system for executing the blockchain-based inpatient management method as described above, the blockchain-based inpatient management system comprising:

[0188] The hospitalization status update module is used to obtain the inpatient registration data, and update the hospitalization status according to the inpatient registration data to obtain the inpatient update data;

[0189] The inpatient resource occupancy processing module is used to process the inpatient resource occupancy according to the inpatient update data to obtain the inpatient resource occupancy data;

[0190] The inpatient resource optimization allocation module is used to dynamically predict resource demand and optimize allocation based on the inpatient resource occupancy data to obtain inpatient resource optimization allocation data;

[0191] The inpatient resource utilization monitoring module is used to implement resource scheduling and monitor resource utilization in real time based on the inpatient resource optimization allocation data, and obtain inpatient resource utilization monitoring data to add to the preset inpatient management blockchain.

[0192] Therefore, from any point of view, the embodiments should be regarded as illustrative and non-restrictive, and the scope of the present invention is limited by the attached application documents rather than the above description, and it is intended that all changes within the meaning and scope of equivalent elements of the application documents are included in the present invention.

[0193] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A method for managing inpatients based on blockchain, characterized in that: The following steps are involved: Step S1: Acquire inpatient registration data, and update the hospitalization status according to the inpatient registration data to obtain inpatient update data; Step S2: Processing the inpatient resource occupancy status according to the inpatient update data to obtain inpatient resource occupancy status data; Step S3: dynamically predict resource demand and optimize allocation based on hospital resource occupancy data to obtain hospital resource optimization allocation data; Step S4: Implement resource scheduling and monitor resource utilization in real time according to the inpatient resource optimization allocation data, and obtain inpatient resource utilization monitoring data to add to the preset inpatient management blockchain; Step S3 is specifically as follows: Extracting characteristics of inpatient resource occupancy according to the inpatient resource occupancy data to obtain characteristic data of inpatient resource occupancy; According to the characteristic data of inpatient resource occupancy, the demand for inpatient resources is predicted to obtain the predicted data of inpatient resource demand; According to the forecast data of inpatient resource demand, an inpatient resource optimization allocation model is constructed to obtain an inpatient resource optimization resource allocation model; Generate an inpatient resource optimization plan based on the inpatient resource optimization resource allocation model to obtain inpatient resource optimization plan data; According to the inpatient resource optimization plan data, the on-chain allocation record is performed to obtain the inpatient resource optimization allocation data; The specific feature extraction of hospitalization resource occupancy is as follows: Perform data repair according to the inpatient resource occupancy data to obtain inpatient resource occupancy repair data; Perform target coding according to the inpatient resource occupancy repair data to obtain inpatient resource occupancy target coding data; According to the target coding data of inpatient resource occupancy, hierarchical feature extraction is performed to obtain the first inpatient resource occupancy characteristic data, wherein the hierarchical feature extraction is that the system performs stratification by resource type according to the coded data, and extracts the key features of each layer. The first inpatient resource occupancy characteristic data includes bed layer characteristic data, medical and nursing layer characteristic data, and equipment layer characteristic data. The bed layer characteristic data is the characteristics of bed occupancy time, idle rate and turnover rate. The medical and nursing layer characteristic data is the workload of doctors and nurses and the frequency of ward rounds. The equipment layer characteristic data is the characteristics of equipment use time, use frequency and waiting time. Cluster calculation is performed based on the target coding data of inpatient resource occupancy to obtain cluster data of inpatient resource occupancy. The cluster calculation is that the system clusters the data of each bed, equipment, and medical resource to form different categories of resource occupancy, which is convenient for identifying the pattern of resource use. Inpatients are divided into three categories: high occupancy group: long bed occupancy time and frequent equipment use; medium occupancy group: moderate bed occupancy time and average equipment use frequency; low occupancy group: low bed and equipment use frequency; Extracting cluster distribution features of the clustered data of hospitalization resource occupancy to obtain second characteristic data of hospitalization resource occupancy, wherein the cluster distribution feature extraction is that the system extracts the distribution features of each cluster according to the clustering results, including the center point of the cluster, the distribution density of the data points in each cluster, and the boundary conditions of the cluster; Perform characteristic density calculation according to the first inpatient resource occupancy characteristic data and the second inpatient resource occupancy characteristic data, to obtain the first inpatient resource occupancy characteristic density data and the second inpatient resource occupancy characteristic density data respectively; Performing feature mutual information analysis based on the first inpatient resource occupancy feature data and the second inpatient resource occupancy feature data to obtain inpatient resource occupancy feature mutual information data; According to the first inpatient resource occupancy characteristic density data, the second inpatient resource occupancy characteristic density data and the inpatient resource occupancy characteristic mutual information data, nonlinear feature fusion is performed on the first inpatient resource occupancy characteristic data and the second inpatient resource occupancy characteristic data to obtain inpatient resource occupancy characteristic data; The construction of the optimal allocation model of hospitalization resources is as follows: Calculate the difference between supply and demand of inpatient resources based on the forecast data of inpatient resource demand, and obtain the difference between supply and demand of inpatient resources; According to the difference data of hospitalization resource supply and demand, the hospitalization resource cost minimization constraint is performed to obtain the hospitalization resource supply optimization data; Prioritize hospitalization resources according to the hospitalization resource supply optimization data to obtain hospitalization resource priority ranking data; Based on the priority ranking data of inpatient resources, simulated allocation and iterative optimization are carried out to build an inpatient resource optimization resource allocation model.

2. The method according to claim 1, characterized in that Step S1 is specifically as follows: Obtain patient identity data, and perform electronic file query based on the patient identity data to obtain the patient's electronic file data; Generate a unique patient identification based on the patient's electronic file data to obtain the patient's unique identification data; The inpatient registration data is obtained by performing on-chain registration records based on the patient's unique identification data; Generate initial hospitalization data based on inpatient registration data to obtain initial hospitalization data; The status update is confirmed based on the initial hospitalization data to obtain updated data of the inpatients.

3. The method according to claim 1, characterized in that Step S2 is specifically as follows: Obtain inpatient resource occupancy data; Analyze the inpatient resource occupancy situation based on the inpatient resource occupancy data and the inpatient update data to obtain the inpatient resource occupancy analysis data; According to the inpatient resource occupancy analysis data, on-chain resource records are made to obtain inpatient resource occupancy data.

4. The method according to claim 1, characterized in that: The characteristic density calculation is as follows: Estimating the distribution type according to the first inpatient resource occupancy characteristic data and the second inpatient resource occupancy characteristic data, and obtaining first characteristic distribution type estimation data and second characteristic distribution type estimation data respectively; Using the first characteristic distribution type estimation data, the first inpatient resource occupancy characteristic data is labeled with a distribution type to obtain first characteristic distribution type labeled data, and using the second characteristic distribution type estimation data, the second inpatient resource occupancy characteristic data is labeled with a distribution type to obtain second characteristic distribution type labeled data; Perform kernel density estimation and distribution fitting based on the first feature distribution type labeled data and the second feature distribution type labeled data to obtain first feature distribution density data and second feature distribution density data; Density calculation is performed based on the first characteristic distribution density data and the second characteristic distribution density data to obtain first inpatient resource occupancy characteristic density data and second inpatient resource occupancy characteristic density data, respectively.

5. The method according to claim 1, characterized in that The specific nonlinear feature fusion is: A density-weighted feature mutual information matrix is ​​generated according to the first inpatient resource occupancy characteristic density data, the second inpatient resource occupancy characteristic density data, and the inpatient resource occupancy characteristic mutual information data to obtain density-weighted feature mutual information matrix data; Constructing a feature interaction network for the first inpatient resource occupancy feature data and the second inpatient resource occupancy feature data according to the density-weighted feature mutual information matrix data to obtain feature interaction network data; A regression model and a neural network are constructed according to the feature interaction network data to obtain a first nonlinear interaction feature model and a second nonlinear interaction feature model respectively; Generate nonlinear interaction features according to the first nonlinear interaction feature model and the second nonlinear interaction feature model to obtain first nonlinear interaction feature data and second nonlinear interaction feature data respectively; Performing nonlinear feature fusion on the first inpatient resource occupancy characteristic data and the second inpatient resource occupancy characteristic data according to the first nonlinear interactive characteristic data to obtain first occupancy nonlinear characteristic data; According to the second nonlinear interactive feature data, nonlinear feature fusion is performed on the first inpatient resource occupancy feature data and the second inpatient resource occupancy feature data to obtain second occupancy nonlinear feature data; The first occupancy situation nonlinear characteristic data and the second occupancy situation nonlinear characteristic data are arranged in layers to obtain the inpatient resource occupancy situation characteristic data.

6. The method according to claim 1, characterized in that Step S4 is specifically as follows: Generate an inpatient resource scheduling plan based on the inpatient resource optimization allocation data to obtain inpatient resource scheduling plan data; According to the inpatient resource scheduling plan data, IoT data is collected and scheduled and adjusted to obtain inpatient resource scheduling collection data; According to the inpatient resource scheduling and collection data, on-chain monitoring records are performed to obtain inpatient resource utilization monitoring data.

7. A blockchain-based inpatient management system, characterized in that: For executing the inpatient management method based on blockchain as claimed in claim 1, the inpatient management system based on blockchain comprises: The hospitalization status update module is used to obtain the inpatient registration data, and update the hospitalization status according to the inpatient registration data to obtain the inpatient update data; The inpatient resource occupancy processing module is used to process the inpatient resource occupancy according to the inpatient update data to obtain the inpatient resource occupancy data; The inpatient resource optimization allocation module is used to dynamically predict resource demand and optimize allocation based on the inpatient resource occupancy data to obtain inpatient resource optimization allocation data; The inpatient resource utilization monitoring module is used to implement resource scheduling and monitor resource utilization in real time based on the inpatient resource optimization allocation data, and obtain inpatient resource utilization monitoring data to add to the preset inpatient management blockchain.

Images