A blood dialysis medical association resource collaborative management system and method based on intelligent bed arrangement decision

Abstract

The application discloses a blood dialysis medical alliance resource collaborative management system and method based on intelligent bed arrangement decision. The system comprises the following steps: acquiring medical order information of dialysis patients; constructing a multi-constraint rule matrix containing rigid constraint conditions and flexible optimization conditions; performing bed matching through a bed arrangement algorithm engine based on the multi-constraint rule matrix and the medical order information, generating an initial bed arrangement plan; listening to home dynamic events and responding to regional referral of the medical alliance, when a dynamic event triggers a treatment demand or the regional referral of the medical alliance proposes a referral application, incrementally rearranging the initial bed arrangement plan based on the multi-constraint rule matrix, generating and outputting a target bed arrangement plan. The application realizes the leap from passive recording to active decision by constructing the multi-constraint rule matrix and the bed arrangement algorithm engine, automatically fuses rigid constraints and flexible optimization, supports incremental rearrangement, greatly improves the bed arrangement accuracy and efficiency, and effectively solves the problems that the existing bed arrangement management relies on manual experience and cannot cope with dynamic changes, resulting in low efficiency and errors.

Description

Technical Field

[0001] This invention relates to the fields of medical information technology and patient management technology, and in particular to a collaborative management system and method for resources of hemodialysis medical consortia based on intelligent bed scheduling decision-making. Background Technology

[0002] Hemodialysis is the core treatment for end-stage renal disease patients. With the extended survival time of dialysis patients and the expansion of medical insurance coverage, large tertiary hospital dialysis centers often have 200-300, or even more, dialysis patients. Faced with such a large patient population, the overall allocation of treatment resources has become a key bottleneck restricting medical quality and safety. Patient bed allocation management is the core of the overall allocation of dialysis treatment resources. Currently, patient bed allocation management has not yet achieved intelligent decision-making and still relies on manual scheduling; patients referred within the region need to undergo infectious disease testing at the receiving hospital and wait for the results before a treatment bed can be arranged; dialysis patient management is still mainly in-hospital management, with discontinuous data from home management, making full-cycle management difficult to implement. The main problems are as follows: Currently, dialysis centers in China face a polarized yet ultimately convergent dilemma in their treatment scheduling management. Some centers remain entirely manual, relying on paper documents or spreadsheets. They manually verify patient information, including infectious disease zoning (isolation and non-isolation areas for hepatitis B, hepatitis C, etc.), treatment modes (HD / HDF / HP), equipment types, and treatment times. Furthermore, the information asymmetry between nurses / patients, doctors / patients, and between doctors and nurses due to changes in patient conditions and needs leads to an exponential increase in scheduling workload. This makes it extremely difficult to avoid errors in treatment scheduling information caused by human oversight, resulting in conflicts between nurses / patients, doctors / patients, and between doctors and nurses, ultimately impacting patient satisfaction.

[0003] The dialysis centers that have already implemented an information management system have technical limitations in their software functionality. As an independent management software separate from the hospital's HIS system, the dialysis center information management system's bed scheduling module has the following defects: First, data silos: Long-term dialysis patients are managed as outpatients, and the dialysis center information management system and the hospital's HIS system are independent of each other. Medical orders need to be manually entered or imported in batches, making real-time synchronization impossible. Second, pseudo-intelligent bed scheduling logic, merely acting as an electronic whiteboard to display the status of bed scheduling information: The existing dialysis center information management system does not have a built-in decision-making algorithm for patient bed scheduling. Scheduling staff still need to manually adjust patient treatment bed scheduling information based on changes in medical orders, patient conditions, and needs. This remains a passive bed scheduling system dominated by human experience, rather than a proactive intelligent matching by the system. It cannot cope with dynamic emergencies such as changes in patient conditions, needs, and equipment malfunctions, essentially not escaping the inefficient manual decision-making model. Furthermore, because the existing information management system's built-in bed scheduling module is bound to patient treatment and nursing information, manual adjustment of bed arrangements is still required, followed by manual entry of patient bed change information according to the system's built-in bed scheduling module logic, which actually increases the workload of patient bed scheduling management.

[0004] Furthermore, during the referral process from county-level hospitals to tertiary dialysis centers, patients face stringent requirements for infectious disease management in hemodialysis. These patients must re-book appointments at the tertiary hospital, undergo infectious disease screening, and wait for test results before the dialysis center can allocate beds for treatment. This prolongs waiting times and limits the utilization rate of dialysis beds. If patients do not wait for infectious disease test results, they can only undergo the more expensive CRRT treatment. This situation not only inconveniences patients seeking dialysis in other locations but also increases their waiting costs and reduces the efficiency of regional medical resource utilization.

[0005] Furthermore, hemodialysis patients are in the uremia stage of chronic renal failure and are a key population for national chronic disease management. Currently, all medical management activities and information management system functions for dialysis patients are limited to the time patients are in the hospital. Once a patient leaves the hospital, the dialysis center cannot obtain their home health data (blood pressure, weight, arteriovenous fistula function, etc.), sudden discomfort, or requests for leave for treatment in a timely manner. The direct consequence of this information gap is that when a patient is unable to come to the hospital for treatment on time due to a sudden situation at home, the dialysis center cannot release the patient's treatment bed and reserve consumables in advance, resulting in the treatment bed being forced to sit idle, as well as the ineffective consumption of dialysis consumables, increasing departmental costs and wasting medical resources. Summary of the Invention

[0006] This invention provides a collaborative management system and method for resources within a hemodialysis medical consortium based on intelligent bed scheduling decision-making. It aims to improve the inefficiency and error-prone nature of existing hemodialysis center bed scheduling management, which relies on manual experience and is unable to cope with dynamic emergencies. Simultaneously, it addresses the problems of delayed referral information, poor coordination, and bed conflicts or vacancies within the medical consortium, affecting the timeliness of patient treatment, as well as the shortcomings of the separation between patient home monitoring data and in-hospital resources, which prevents the response to sudden acute events at home.

[0007] To achieve the above-mentioned objectives, the following technical solution is provided: Provides a collaborative management system for hemodialysis medical consortia based on intelligent bed scheduling decisions, including: The intelligent decision-making bed scheduling management subsystem is used to acquire the medical order information of dialysis patients, construct a multi-constraint rule matrix containing rigid constraints and flexible optimization conditions, and perform joint matching of beds and dialysis equipment through the bed scheduling algorithm engine based on the multi-constraint rule matrix and the medical order information to generate an initial bed scheduling plan. It also listens for dynamic events, and when the dynamic event is triggered, it incrementally rearranges the initial bed scheduling plan based on the multi-constraint rule matrix to generate and output the target bed scheduling plan. The medical consortium referral collaboration platform is used to receive referral application information sent by primary healthcare institutions, trigger the pre-test process to obtain infectious disease test results, and then write the bed allocation algorithm engine triggered by the doctor's order to pre-allocate beds according to available beds. After receiving the hospital's confirmation instruction, the pre-allocated beds are switched to a locked state. The home monitoring and early warning linkage subsystem is used to collect home monitoring data uploaded by patient terminals, compare the home monitoring data with preset thresholds to determine whether an abnormal early warning is triggered, and when a critical value early warning is triggered, the bed scheduling algorithm engine is activated to arrange an expedited bed. When an intention to request leave is detected, the bed scheduling algorithm engine is activated to release a bed.

[0008] Furthermore, it also includes: The system includes a management terminal, a nurse terminal, a doctor terminal, and a patient terminal, all of which are communicatively connected to the intelligent bed scheduling management subsystem. The management terminal is used for system configuration, importing and exporting medical orders, setting bed scheduling rules, verifying and adjusting bed scheduling, managing patients referred by medical consortia, managing home monitoring data, and viewing data statistics; The nurse terminal is used to view patient bed assignment information, mark patients on leave, make temporary adjustments to bed assignments for the day, and mark blood collection information; The doctor's terminal is used to enter medical orders, view patient bed arrangement information, view home monitoring data and abnormal alerts, approve patient leave applications and medical consortium referral applications, and view infectious disease test results. The patient terminal is used to view personal bed arrangements, upload home health data, submit changes in treatment needs or leave applications, and view home health education content.

[0009] It also provides a collaborative management method for hemodialysis medical consortium resources based on intelligent bed scheduling decisions, including: Obtain medical orders from dialysis patients; Construct a multi-constraint rule matrix that includes both rigid constraints and flexible optimization conditions; Based on the multi-constraint rule matrix and the medical order information, the bed allocation algorithm engine performs joint matching of beds and dialysis equipment to generate an initial bed allocation plan. Listen for dynamic events. When a treatment need is triggered by a dynamic event, incrementally rearrange the initial bed scheduling plan based on the multi-constraint rule matrix to generate and output the target bed scheduling plan.

[0010] Furthermore, the rigid constraints include infectious disease zoning and isolation constraints, treatment mode and equipment function compatibility constraints, and time period capacity constraints. Based on the multi-constraint rule matrix and the medical order information, the bed allocation algorithm engine performs joint matching of beds and dialysis equipment to generate an initial bed allocation plan, including: Based on the infectious disease identifier in the medical order information, the corresponding isolation zone is matched; Verify whether the number of patients admitted to each isolation zone during the target treatment period exceeds the corresponding bed capacity limit; Based on the dialysis treatment parameters in the medical order information, dialysis devices that support the corresponding dialysis treatment mode and are in normal condition are selected within the matched isolation zone, and the physical beds corresponding to the selected dialysis devices are allocated to the dialysis patients.

[0011] Furthermore, the flexible optimization conditions include scheduling strategy optimization conditions and treatment period and bed continuity preference conditions; The step of generating an initial bed allocation plan by performing joint matching of beds and dialysis equipment through a bed allocation algorithm engine based on the multi-constraint rule matrix and the medical order information also includes: When the medical order information includes a hemodialysis filtration or hemoperfusion treatment mode, the treatment mode is fixed on a fixed date every week according to a periodic scheduling strategy, or the number of days between each treatment of the treatment mode is evenly distributed according to a scheduling strategy that evenly distributes the number of treatments in the month. Provided that the rigid constraints are met, dialysis patients are preferentially assigned to their historically preferred beds and treatment times.

[0012] Furthermore, the dynamic events include medical order change events, patient leave events, equipment failure events, and patient status change events; When the dynamic event treatment demand is triggered, the initial bed scheduling plan is incrementally rearranged based on the multi-constraint rule matrix to generate and output a target bed scheduling plan, including: When the medical order change event is detected, the original bed allocation plan for the corresponding patient is withdrawn, and the bed allocation is re-executed according to the changed medical order information; When the patient's leave event is detected, the target bed for the target time period allocated to the corresponding patient during the leave period is marked as released. When the equipment failure event is detected, the corresponding dialysis patient originally assigned to the faulty equipment is transferred to an available bed that meets the rigid constraint conditions; When the patient status change event is detected, the subsequent bed scheduling plan for the corresponding patient is cleared.

[0013] Furthermore, it also includes the collaborative scheduling process for referrals within medical consortia: Receive referral requests from primary healthcare institutions; After triggering the pre-detection process and obtaining the infectious disease test results, the results are written into the bed scheduling algorithm engine through medical orders; Pre-allocation is performed based on available beds, and the pre-allocation results are pushed to the patient's terminal; Upon receiving confirmation from the hospital, the pre-assigned bed will be switched to a locked state.

[0014] Furthermore, the process of triggering the pre-detection procedure, obtaining the infectious disease test results, and then writing the medical order into the bed scheduling algorithm engine includes: If test results are mutually recognized within the region, the partition identifier of recent infectious disease test records can be directly extracted. After the doctor confirms, the medical order is written into the bed scheduling algorithm engine. If test results are not mutually recognized within the region, the medical order is connected to the outpatient system and a test order is generated. After receiving the qualified infectious disease test results from the retest, the medical order is written into the bed scheduling algorithm engine.

[0015] Furthermore, it also includes a home monitoring and bed allocation linkage process: Collect home monitoring data uploaded by patients' terminals; The home monitoring data is compared with personalized thresholds to determine whether an abnormal warning has been triggered. When a critical value warning is triggered, the patient's needs will be coordinated with the bed scheduling algorithm engine to arrange an expedited bed; When a leave request is detected, the bed allocation algorithm engine is activated to release a bed.

[0016] Furthermore, it also includes multi-terminal role collaboration and statistical backtracking processes: Based on the role permissions of management terminals, nurse terminals, doctor terminals, and patient terminals, bed allocation information is pushed to specific users. Record bed utilization rate, equipment usage data, patient change data and consumable usage data, generate multi-dimensional statistical backtracking charts, and trigger dialysis indicator monitoring reminders according to preset cycles, and push the list of patients who should have blood drawn for monitoring to each terminal. Beneficial effects:

[0017] This invention imports medical order information and, through medical order recognition, achieves intelligent zoning, bed allocation, and equipment matching for patient treatment based on these orders. It overcomes the limitations of traditional dialysis center information management systems where the bed allocation module only displays information and lacks automatic matching and decision-making capabilities. This invention uses medical order change events as driving signals, employing multi-dimensional constraints for intelligent bed allocation, and supports incremental rescheduling under dynamic emergencies. It represents a leap from passive record-based bed allocation to proactive decision-making bed allocation, and also supports switching between two special modes: weekly and date-based bed allocation, to adapt to the management habits of different centers.

[0018] This invention utilizes a medical consortium collaborative platform to move infectious disease testing to the referral application stage. Test results and treatment orders automatically trigger bed allocation, enabling a reverse scheduling model where referred patients can apply for treatment in different locations, undergo testing in advance, be allocated beds early, and be immediately admitted to the hospital upon arrival. This fundamentally solves the structural waste of resources caused by testing first and then allocating beds later. It also overcomes the problem of excessively long traditional processes, where patients have to wait for hours or even days.

[0019] This invention collects patients' vital signs data at home via smart terminals and leave requests submitted through a patient-side mini-program. When abnormal data indicates a treatment need or a patient proactively requests leave, the system automatically triggers bed allocation adjustments or bed release commands, achieving a fully automated closed-loop management system encompassing home status, early warning triggering, bed allocation linkage, and resource release / redistribution. This overcomes the problem of traditional dialysis patient management ending at the hospital and the lack of home-based health information. Attached Figure Description

[0020] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a schematic diagram of the structure of a collaborative management system for hemodialysis medical consortium based on intelligent bed scheduling decision-making, which is a specific implementation method.

[0022] Figure 2This is a flowchart illustrating a method for collaborative resource management within a hemodialysis medical consortium based on intelligent bed scheduling decisions, implemented in a specific manner.

[0023] Figure label: Intelligent bed scheduling management subsystem; 101. Management terminal; 102. Nurse terminal; 103. Doctor terminal; 104. Patient terminal; 20. Medical consortium referral collaboration platform; 30. Home monitoring and early warning linkage subsystem. Detailed Implementation

[0024] To make the technical problems solved, the technical solutions, and the beneficial effects of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings of the embodiments. It should be understood that the specific embodiments described herein are merely illustrative and are not intended to limit the invention.

[0025] Hemodialysis is a core treatment for end-stage renal disease patients. With the extended survival time of dialysis patients and expanded medical insurance coverage, large tertiary hospital dialysis centers have a massive patient population. Faced with this large patient group, the coordinated allocation of treatment resources has become a key bottleneck restricting medical quality and safety. Currently, patient bed scheduling management has not yet achieved intelligent decision-making and still relies on manual scheduling. Existing systems merely function as electronic whiteboards; the bed scheduling logic lacks decision-making algorithms and still relies on human experience, making it unable to cope with dynamic emergencies. Furthermore, system constraints increase the workload of manual adjustments. In addition, patients referred within the region need to undergo infectious disease testing at the receiving hospital and wait for the results before being assigned a treatment bed, leading to a structural waste of resources. Dialysis patient management remains primarily in-hospital, with discontinuous data from home management, making full-cycle management difficult to implement.

[0026] Some of the terms or nouns that appear in the description of the embodiments of the present invention shall be interpreted as follows: HIS: Hospital Information System.

[0027] CRRT: Continuous Renal Replacement Therapy.

[0028] HD: Hemodialysis.

[0029] HDF: Hemodiafiltration.

[0030] HP: Hemoperfusion.

[0031] AVF: Arteriovenous Fistula.

[0032] LLM: Large Language Model.

[0033] The following will further elaborate on the resource collaborative management system and method for hemodialysis medical consortia based on intelligent bed scheduling decision-making, with reference to specific embodiments and accompanying drawings. Example 1

[0034] Please see Figure 1 The hemodialysis medical consortium resource collaborative management system based on intelligent bed scheduling decision disclosed in this embodiment includes: The intelligent bed scheduling management subsystem 10 is used to acquire the medical order information of dialysis patients, construct a multi-constraint rule matrix containing rigid constraints and flexible optimization conditions, and perform joint matching of beds and dialysis equipment through the bed scheduling algorithm engine based on the multi-constraint rule matrix and the medical order information to generate an initial bed scheduling plan. It also listens for dynamic events, and when the dynamic event triggers a treatment demand, it incrementally rearranges the initial bed scheduling plan based on the multi-constraint rule matrix to generate and output the target bed scheduling plan.

[0035] The medical consortium referral collaboration platform 20 is used to receive referral application information sent by primary healthcare institutions, trigger the pre-test process to obtain infectious disease test results, write medical orders into the bed allocation algorithm engine, pre-allocate beds according to available beds, and switch the pre-allocated beds to a locked state after receiving the hospital confirmation instruction.

[0036] The home monitoring and early warning linkage subsystem 30 is used to collect home monitoring data uploaded by the patient terminal, compare the home monitoring data with a preset threshold to determine whether an abnormal early warning is triggered, and when a critical value early warning treatment demand is triggered, the bed scheduling algorithm engine is linked to arrange an expedited bed. When an intention to request leave is detected, the bed scheduling algorithm engine is linked to release a bed.

[0037] In this embodiment, it also includes: management terminal 101, nurse terminal 102, doctor terminal 103 and patient terminal 104, each terminal being communicatively connected to the intelligent bed scheduling management subsystem, the medical consortium referral collaboration platform and the home monitoring and early warning linkage subsystem respectively; The management terminal is used for system configuration, importing and exporting medical orders, setting bed scheduling rules, verifying and adjusting bed scheduling, managing patients referred by medical consortia, managing home monitoring data, and viewing data statistics; The nurse terminal is used to view patient bed assignment information, mark patients on leave, make temporary adjustments to bed assignments for the day, and mark blood collection information; The doctor's terminal is used to enter medical orders, view patient bed arrangement information, view home monitoring data and abnormal alerts, approve patient leave applications, approve medical consortium referral applications, and issue infectious disease testing orders. The patient terminal is used to view personal bed arrangements, upload home health data, submit changes in treatment needs or leave applications, and view home health education content.

[0038] The aforementioned system constructs an intelligent bed scheduling algorithm engine driven by medical orders, automatically integrating zoning and isolation rules, instrument function adaptation, time constraints, and changes in medical order information, replacing traditional manual experience-based decision-making. It breaks down barriers within the medical consortium, prioritizing infectious disease testing orders for dialysis referral patients, facilitating priority review channels, and enabling mutual recognition of infectious disease test results. This allows for a reverse scheduling model where patients referred from primary care facilities can apply for treatment in other locations, undergo testing in advance, be scheduled for beds early, and start dialysis immediately upon arrival, significantly shortening waiting times for referrals. Furthermore, it establishes a real-time linkage and early warning mechanism between home monitoring and in-hospital treatment bed scheduling systems, breaking down information barriers that prevent patients from notifying dialysis centers in advance due to unforeseen circumstances. This automatically triggers instructions to adjust treatment times and beds or release treatment resources, eliminating hidden waste of bed resources and dialysis consumables at the source. This invention comprehensively optimizes the medical experience within regional medical consortia and maximizes the efficiency of dialysis center resource utilization. Example 2

[0039] This invention provides a method for collaborative management of resources within a hemodialysis medical consortium based on intelligent bed scheduling decision-making. Please refer to [link / reference]. Figure 2 The system acquires the medical orders of dialysis patients; constructs a multi-constraint rule matrix containing rigid constraints and flexible optimization conditions; based on the multi-constraint rule matrix and medical order information, it performs bed matching through a bed allocation algorithm engine to generate an initial bed allocation plan; it listens for dynamic events, and when a dynamic event triggers a treatment demand, it incrementally rearranges the initial bed allocation plan based on the multi-constraint rule matrix to generate and output the target bed allocation plan.

[0040] Specifically, this method relies on a smart bed scheduling management subsystem, which operates independently of the hospital's HIS system and uses medical orders as the sole reliable data source. The overall system architecture comprises three main modules: smart bed scheduling, home management, and integrated management within the medical consortium. Medical order information includes basic patient information, infectious disease markers (such as all negative, monoclonal antibody positive, dual antibody positive, hepatitis B, syphilis, hepatitis C), dialysis treatment parameters (such as HD / HDF / HP, frequency, duration), patient status, and dialysis day selection. A multi-constraint rule matrix structures the complex bed scheduling rules, and the bed scheduling algorithm engine automatically matches beds based on this matrix.

[0041] By constructing a multi-constraint rule matrix, the complex bed scheduling rules are structured and algorithmized, replacing the traditional manual experience-based and passive electronic whiteboard recording methods. The bed scheduling algorithm engine can automatically integrate rigid constraints and flexible optimization to quickly calculate the optimal initial bed scheduling plan. Simultaneously, by monitoring dynamic events and triggering incremental rescheduling, the system can respond in real time to unexpected situations such as changes in medical orders and equipment failures, avoiding the increased workload and error-prone issues associated with manual adjustments. This represents a leap from passive recording to proactive decision-making, significantly improving the accuracy and efficiency of bed scheduling.

[0042] As an example, in a dialysis center with 250 patients, after the system initializes and configures 60 bed resources and imports the medical orders of 250 patients, the algorithm engine can instantly complete the full scheduling. The administrator only needs to check the highlighted special cases, which greatly improves management efficiency.

[0043] In some embodiments, rigid constraints include infectious disease zoning isolation constraints, treatment mode adaptation constraints, and time-slot capacity constraints. Based on the multi-constraint rule matrix and medical order information, a bed matching algorithm engine is used to generate an initial bed matching plan, including: matching isolation zones according to the infectious disease identifier in the medical order information and allocating physical beds; screening dialysis equipment that supports the corresponding dialysis treatment mode and is in normal condition according to the dialysis treatment parameters in the medical order information, and binding the dialysis equipment to the physical beds; and verifying whether the number of patients admitted in each treatment time slot exceeds the bed capacity limit.

[0044] Specifically, the infectious disease zoning and isolation constraints ensure that hepatitis B-positive patients are only assigned to the hepatitis B isolation area; the treatment mode adaptation constraints ensure that HD mode is matched with HD machines in the negative area, and HDF mode is matched with HDF machines; and the time-slot capacity constraints ensure that the number of patients admitted during the morning, afternoon, and evening shifts does not exceed the physical bed capacity limit. Through strict infectious disease zoning and isolation constraints, nosocomial cross-infection is effectively prevented. The adaptation constraints between treatment modes and equipment functions ensure the smooth implementation of treatment. The time-slot capacity constraints avoid overloading of shifts, ensuring the rational allocation of medical resources and medical safety.

[0045] In some embodiments, the flexible optimization conditions include scheduling strategy optimization conditions and bed continuity preference conditions; based on the multi-constraint rule matrix and medical order information, bed matching is performed through the bed scheduling algorithm engine to generate an initial bed scheduling plan, which also includes: when the medical order information contains a specific dialysis treatment mode, the specific dialysis treatment mode is kept constant on a preset date according to the periodic scheduling strategy, or the interval days of the specific dialysis treatment mode are evenly distributed according to the monthly treatment frequency equal distribution scheduling strategy; under the premise of meeting the rigid constraint conditions, dialysis patients are preferentially assigned to historically preferred beds and historically preferred treatment periods.

[0046] Specifically, a weekly scheduling strategy ensures that special treatment modalities (such as HP and HDF) are scheduled for a fixed day each week (avoiding Mondays and Tuesdays to circumvent health insurance issues during the alternation of months with varying numbers of days in the calendar); a scheduling strategy that distributes treatments evenly across the month ensures that the intervals between treatments for special modalities are as equal as possible within a given month. Through optimized scheduling strategies, the treatment intervals for special modalities become more scientifically sound and compliant with medical treatment guidelines. Furthermore, by prioritizing bed continuity preferences and allocating historically preferred beds and time slots, the patient experience and treatment comfort are improved, while simultaneously reducing the time nurses spend familiarizing themselves with different patient habits, thus increasing overall nursing efficiency.

[0047] Please see Figure 1 In some embodiments, dynamic events include medical order change events, patient leave events, equipment failure events, and patient status change events. When a dynamic event is triggered, the initial bed scheduling plan is incrementally rearranged based on a multi-constraint rule matrix to generate and output a target bed scheduling plan, including: when a medical order change event is detected, withdrawing the initial bed scheduling plan and re-executing the bed scheduling according to the changed medical order information; when a patient leave event is detected, marking the target bed as released; when an equipment failure event is detected, transferring the corresponding dialysis patient to an available bed; and when a patient status change event is detected, clearing the subsequent bed scheduling plan.

[0048] By setting explicit incremental rescheduling strategies for different types of dynamic events, the system can make rapid adjustments, either locally or in a targeted manner, without overturning the overall plan. For example, when a patient requests leave, only the corresponding bed is released; when equipment malfunctions, the patient is automatically transferred, thus ensuring the continuity and stability of the bed allocation plan and greatly reducing the impact of emergencies on the overall treatment schedule.

[0049] In some embodiments, when a device malfunction event is detected, the corresponding dialysis patient is transferred to an available bed, including: receiving malfunction information marked by a nurse terminal and updating the device status of the dialysis device; searching for available beds in the same partition that support the corresponding dialysis treatment mode under the constraints of a multi-constraint rule matrix; shifting the initial bed allocation plan to an available bed and generating an adjustment notification to push to the corresponding terminal.

[0050] By receiving fault information from nurse terminals and updating equipment status in real time, the bed scheduling algorithm engine ensures it has the latest equipment availability. Under the constraints of the same zone and supporting corresponding treatment modes, it searches for available beds and moves plans accordingly, ensuring both the safety baseline for infectious disease isolation and maintaining the continuity of treatment modes, achieving seamless transitions in case of failure.

[0051] Please see Figure 1In some embodiments, the method further includes a referral coordination and scheduling process within the medical consortium: receiving referral application information sent by a primary healthcare institution; triggering a pre-testing process to obtain qualified infectious disease test results, and after doctor confirmation, writing the medical order into the bed allocation algorithm engine; pre-allocating beds based on available beds and pushing the pre-allocation results to the patient's terminal; and after receiving a hospital confirmation instruction, switching the bed status corresponding to the pre-allocation result to a locked state.

[0052] By establishing a collaborative referral scheduling process within the medical consortium, the cumbersome traditional process of testing first and then scheduling has been broken. Through pre-testing procedures and bed pre-allocation, it is possible to apply from other locations, conduct tests in advance, schedule beds in advance, and start treatment immediately upon arrival. This reduces referral waiting time to minutes, avoids patients being forced to choose expensive CRRT treatment due to waiting, optimizes the allocation of regional medical resources, and improves bed utilization.

[0053] In some embodiments, a pre-testing process is triggered to obtain qualified infectious disease test results. After the doctor confirms the results, the medical order is written into the bed scheduling algorithm engine. This includes: if the test results are mutually recognized within the region, the partition identifier of the recent test record is directly extracted and the medical order is written into the bed scheduling algorithm engine; if the test results are not mutually recognized within the region, a test medical order is generated and connected to the outpatient system to receive the qualified infectious disease test results from the retest. After the doctor confirms the results, the medical order is written into the bed scheduling algorithm engine.

[0054] By differentiating between test results within the region based on whether they are mutually recognized, a differentiated processing strategy was adopted. For mutually recognized results, the medical orders were directly identified; for non-mutually recognized results, the medical orders were routed to the outpatient system for retesting. This ensured the rigor of infectious disease prevention and control while minimizing unnecessary duplicate testing and improving referral efficiency. As an example, a county-level hospital submitted a referral request for a patient who had been on dialysis for 3 years and tested positive for hepatitis C 3 months prior. A doctor at a tertiary hospital prescribed retesting because the patient was older than 3 months. After the patient tested positive for hepatitis C RNA, the system locked onto a pre-assigned bed in the hepatitis C isolation area. Upon arrival at the hospital, the nurse confirmed the patient's arrival, and the patient could be put on dialysis within 15 minutes.

[0055] In some embodiments, the method further includes a home monitoring and bed allocation linkage process: collecting home monitoring data uploaded by the patient's terminal; comparing the home monitoring data with personalized thresholds to determine whether an abnormal warning is triggered; when a critical value warning is triggered, the patient's treatment needs are linked to the bed allocation algorithm engine to arrange expedited beds; when a leave of absence intention is triggered, the bed allocation algorithm engine is linked to release beds.

[0056] Specifically, home monitoring data includes blood pressure, weight, and AVF function. Personalized thresholds can be set based on the patient's dry weight. For example, for a dry weight of 60kg, the weight fluctuation threshold can be set to ±2kg, and the systolic blood pressure threshold can be set to <90 or >180. As an example, the personalized threshold for weight fluctuation can be ±1kg, ±1.5kg, ±2kg, ±2.5kg, or ±3kg, or any value within the range determined by any two of the above values; the warning threshold for systolic blood pressure can be <80, <85, <90, >170, >175, >180, or >190, etc. By collecting home monitoring data and comparing it with personalized thresholds, a data loop between the hospital and outpatient settings is established. When a critical value warning is triggered, the patient's treatment needs are automatically triggered to arrange an expedited bed; when a leave request is triggered, the bed is automatically released. This achieves monitoring upon discharge, warning of abnormalities, and linkage for leave requests, effectively reducing bed vacancy rates and ineffective waste of consumables, while ensuring patient safety.

[0057] In some embodiments, collecting home monitoring data uploaded by the patient terminal includes: receiving voice data, text data, and numerical data uploaded by the patient terminal to form multimodal data; comparing the home monitoring data with personalized thresholds to determine whether an abnormal warning is triggered, including: calling a preset language model to perform intent recognition and urgency analysis on the multimodal data; dynamically adjusting the trigger boundary of the personalized thresholds based on the output of the preset language model, and determining whether an abnormal warning is triggered based on the adjusted trigger boundary. The preset language model can be a medical large language model, a general large language model, or a specialized fine-tuning model, etc., and this embodiment of the invention does not limit this.

[0058] By introducing a pre-defined language model to process multimodal data such as speech, text, and numerical values, the limitations of traditional home monitoring, which relies solely on a single numerical threshold, are overcome. The model can accurately identify the patient's true intentions and the degree of urgency, and dynamically adjust the trigger boundaries, making the early warning mechanism more intelligent and human-centered, and reducing false alarms and missed alarms.

[0059] Please see Figure 1 In some embodiments, a preset language model is invoked to perform intent recognition and urgency analysis on multimodal data, including: converting speech data into speech-text data, and concatenating the speech-text data with text data and numerical data to form a multimodal input sequence; inputting the multimodal input sequence into a fine-tuned medical language model to identify high-risk intent or routine leave request intent; when a high-risk intent is identified, lowering the physiological indicator warning threshold to trigger a first-level warning.

[0060] By converting speech data into text and concatenating it with other data to form a multimodal input sequence, the system fully leverages the specialized understanding capabilities of a finely tuned medical language model within the medical field. When high-risk intent is detected, the system proactively lowers the warning threshold for physiological indicators, achieving a higher level of alert triggering and ensuring timely intervention for potential high-risk conditions such as early signs of heart failure. For example, a patient uploaded a voice message stating they felt short of breath and had gained 2 pounds. The model identified this as an intent to indicate early signs of heart failure and determined it to be high-risk. The system then dynamically lowered the threshold based on the weight gain, triggering a red alert. The patient's treatment needs were directly linked to the bed scheduling system, arranging for emergency diaphragm treatment.

[0061] Please see Figure 1 In some embodiments, obtaining the medical order information of dialysis patients includes: obtaining the medical order information by importing a preset data template or manually entering the information; the method also includes: when the patient status of the dialysis patient is marked as transferred out, discharged, dead or kidney transplanted, clearing the subsequent bed scheduling plan and including it in the change statistics.

[0062] Specifically, the preset data templates support Excel format, with fields including patient name, age, card number, mobile phone number, infectious disease status, patient type, dialysis day selection, treatment mode, and frequency. By supporting both preset data template import and manual entry, the flexibility and efficiency of obtaining medical order information are improved. When a patient's condition undergoes a fundamental change, subsequent plans are automatically cleared and included in the statistics, avoiding unnecessary bed scheduling that occupies resources and ensuring the real-time nature and accuracy of bed scheduling data.

[0063] In some embodiments, the method further includes a multi-terminal role collaboration and statistical backtracking process: based on the role permissions of the management terminal, nurse terminal, doctor terminal and patient terminal, targeted bed allocation information is pushed; bed allocation utilization rate, equipment usage data, patient change data and consumable usage data are recorded, and multi-dimensional statistical backtracking charts are generated.

[0064] Specifically, the management terminal (computer terminal) is responsible for system configuration, importing and exporting medical orders, and setting bed allocation rules; the nurse terminal (mini-program) is responsible for viewing beds, marking leave requests, and making emergency adjustments; the doctor terminal (mini-program) is responsible for approving leave requests and adjusting medical orders; and the patient terminal (mini-program) is responsible for applying for leave and viewing arrangements. Through multi-terminal role collaboration, targeted information delivery and access control are achieved, reducing communication costs among medical staff, nurses, and patients. By recording multi-dimensional data and generating statistical backtesting charts (such as bar charts), solid data support is provided for the dialysis center's operation management, equipment maintenance, and consumable procurement, assisting management in making informed decisions.

[0065] In some embodiments, generating and outputting a target bed scheduling plan includes: generating a bed scheduling calendar view on the management terminal that supports scrolling and annotation; modifying the beds or treatment periods in the target bed scheduling plan in response to manual drag-and-drop adjustment commands; verifying in real time whether the modified target bed scheduling plan meets the rigid constraints, and if the verification passes, recording the operation log and issuing an update notification.

[0066] By generating a bed scheduling calendar view that supports weekly / monthly scrolling and annotation, an intuitive global bed scheduling perspective is provided. Manual drag-and-drop adjustments and real-time verification of rigid constraints are supported, preserving the flexibility of manual intervention while ensuring medical safety through system verification. This guarantees that the final published bed scheduling plan meets actual needs and is absolutely safe and compliant. Upon confirmation, the system automatically pushes update notifications to all mini-programs, completing the information loop.

[0067] To further illustrate the inventive aspects of this invention, the following examples are provided, which are not intended to be limiting. Example 1

[0068] This example uses a dialysis center of a tertiary hospital (with 250 dialysis patients, 60 dialysis machines, divided into morning and afternoon shifts, and 8 beds in the hepatitis B isolation area, 6 beds in the hepatitis C isolation area, and 46 beds in the negative area) to illustrate the specific implementation of the intelligent bed scheduling management subsystem.

[0069] Step 1: System Deployment and Initialization The intelligent bed scheduling management subsystem software is deployed on the local server of the dialysis center. Management and nurse PC terminals are configured at the nurse stations, and nurse station sub-terminals are configured in each treatment area. The system administrator uploads the hospital logo and configures account management (patient accounts, nurse accounts, doctor accounts, and administrator accounts) in the system settings module. Bed resource information is imported in batches in the bed resource maintenance function. Each bed must be labeled with: bed number, its physical area (Hepatitis B isolation area / Hepatitis C isolation area / negative area), the model of the dialysis machine and supported treatment mode (HD / HDF / HP), and its current operating status (normal / faulty / under maintenance).

[0070] Step Two: Medical Order Entry and Management The bed scheduling administrator downloads an Excel template using the batch import function for dialysis orders. The template fields include: patient name, age, card number, mobile phone number, date of birth, infectious disease result (dropdown options: all negative / Hepatitis B / Hepatitis C), patient type (long-term outpatient / inpatient), dialysis day selection (multiple selections from Monday to Saturday), treatment mode (HD / HDF / HP), frequency (e.g., 3 times per week), single session duration (e.g., 4 hours), and patient status (on dialysis / transferred / discharged / dead / kidney transplant). After entering the order information for all 250 dialysis patients in the center into the Excel file according to the template format, the administrator imports it into the system with one click.

[0071] For newly admitted patients or patients requiring changes to their medical orders, administrators can click "Add Medical Order" in the medical order list interface, enter the patient's basic information, dialysis treatment parameters, and patient status one by one, and click "Save" to make the changes effective. The medical order change query function automatically records all additions, modifications, and deletions, and supports querying and exporting by time range.

[0072] Step 3: Automatic matching of bed arrangement algorithm After the bed scheduling administrator selects the bed scheduling cycle (this month / next month) and scheduling mode (weekly scheduling / date scheduling) on ​​the start scheduling interface, they click "Start Scheduling," and the system activates the automatic bed scheduling algorithm. Taking a patient who is completely negative, with a treatment mode of HD (hemodialysis) and frequency of 3 times per week, and HDF (dialysis for the second time per month), and whose dialysis days are Monday, Wednesday, or Friday as an example: 1. Rigid constraint check: Infectious Disease Zoning: If the infectious disease result is negative for all 46 beds, the patient can be assigned to any one of the 46 beds in the negative zone (assuming there are 38 machines supporting HD and 8 machines supporting HDF in the negative zone); if the infectious disease result is positive for Hepatitis B, matching can only be performed within the 8 beds in the Hepatitis B isolation zone. This patient is negative for infectious diseases, and both HD and HDF treatments will be performed in the negative zone with matching treatment equipment.

[0073] Treatment Modes - Instrument Compatibility: For infectious diseases with a negative result, use HD mode and select a dialysis machine with HD function within the negative screening area for matching, 3 times a week; for HDF mode, select a dialysis machine with HDF function within the negative screening area for matching, 2 times a month.

[0074] Time slot capacity: Based on the patient's preferred treatment time slot and the current available capacity of each time slot, the patient's preferred treatment time slot is prioritized, such as the morning. Flexible optimization calculation: If the schedule is weekly: the patient will be scheduled for HD treatment on Monday, Wednesday and Friday mornings. HDF treatment will be scheduled for the first and third Wednesdays of each month (intervals are multiples of 7, with a fifth week in longer months). Tuesdays, Thursdays, Saturdays and Sundays will not be scheduled. If the scheduling is based on the number of treatments per month: the patient will be scheduled for HD treatment on Monday, Wednesday, and Friday mornings. For HDF treatment, the interval between two HDF treatments will be as equal as possible based on the number of treatments per month (e.g., if the month has 31 days and HDF is performed twice a month, the interval between two treatments will be approximately 15-16 days). The system will automatically calculate the optimal scheduling plan, and no treatments will be scheduled on Tuesdays, Thursdays, Saturdays, and Sundays.

[0075] Flexible adjustment of treatment time slots: If all beds for the preferred treatment time slot of a newly admitted patient are already booked, the system will automatically or manually match available beds for other treatment time slots to ensure patient treatment and bed utilization.

[0076] Bed allocation plan generation: Based on the medical orders of all 250 patients, the system completes the bed-time allocation one by one and generates a complete monthly bed allocation calendar.

[0077] Step 4: Manual verification and adjustment The bed allocation administrator enters the bed allocation verification and adjustment interface, and the system automatically displays verification reminders—highlighting special situations encountered during the bed allocation process (such as insufficient beds at a certain time, unscheduled beds for a certain patient, changes in a patient's medical orders, etc.). Administrators can verify bed allocation information daily through the bed allocation display function (supporting scrolling viewing of last week / this week / next week, last month / this month / next month). If a patient strongly requests to be moved to a bed during a specific treatment time slot, such as in the afternoon, the administrator clicks on the patient's cell and selects manual editing. They then select the target bed from the list of available beds for the current time slot. The system automatically checks rigid constraints (whether zoning isolation and instrument compatibility requirements are met). Once approved, the adjustment is confirmed, and the operation record is written to the operation log.

[0078] Step 5: Posting and Notification of Bed Allocation After verification, click the "Verification Complete" button, then click "Publish Schedule." The system will then push the final bed arrangement plan to the following terminals: The nurse app displays a complete bed allocation calendar, allowing nurses to view the bed-patient correspondence for different time slots on the day. Doctor's mini-program: Displays the bed arrangement for all patients; The patient-side mini-program allows users to view their personal treatment bed allocation and dates.

[0079] Step Six: Dynamic Event Triggered Reordering Scenario 1 (Patient Leave Request): A patient submits a leave request for treatment next Wednesday via the patient-side mini-program (stating the reason: family matters), requesting that Wednesday's treatment be rescheduled to Thursday. The system automatically marks the patient's bed for next Wednesday as released and adds the bed to the available bed pool for new patients needing treatment on Wednesday; the incremental rearrangement algorithm is activated, scheduling the patient for the corresponding treatment time slot with available beds on Thursday.

[0080] Scenario 2 (Medical Order Change): A patient's medical order is changed from HDF (twice a month) to HDF (four times a month). The administrator modifies the patient's treatment mode parameters in the medical order list and saves it. The medical order change query module captures this change event and automatically triggers incremental rescheduling—canceling all original bed scheduling plans for the patient, rematching beds that support HDF according to the new medical order (HDF mode change), regenerating subsequent schedules, and pushing medical order change reminders to the nurse's and doctor's mini-programs for review.

[0081] Scenario 3 (Instrument Failure): When a nurse marks the status of a dialysis machine corresponding to a certain bed as faulty in the instrument operation module, the system automatically transfers patients scheduled for that bed that day and subsequent days to other available beds that meet the rigid constraints, and pushes adjustment notifications to the nurse's, doctor's, and patient's mini-programs.

[0082] Step 7: Statistical Backtracking Administrators access the bed scheduling management statistics dashboard to view the following data: Bed occupancy statistics: This month's bed utilization rate is 92%, with saturation rates of 95% in the morning and 88% in the afternoon for each shift. Instrument usage: Ranking of monthly operating time and failure frequency of 60 instruments; Patient changes: This month, there were 243 patients undergoing dialysis, 5 hospitalized, 3 discharged, 1 death, 2 transferred out, 1 kidney transplant recipient, and 8 patients on leave. Consumable usage statistics: This month, XX sets of HD mode consumables, XX sets of HDF mode consumables, and XX sets of HP mode consumables; Statistics on changes in medical orders and bed allocation: There were 15 changes in medical orders and 23 bed adjustments this month (including leave requests, malfunctions, etc.), and the trend of these changes is presented in a bar chart. Dialysis indicator monitoring reminder: The system marks the list of patients who should have blood drawn for monitoring this month (according to the center's quarterly unified arrangement: 3-month cycle), and the reminder label is displayed simultaneously on the nurse's, doctor's, and management terminals. Example 2

[0083] This example demonstrates the specific implementation of the medical consortium referral collaboration platform by referring a hepatitis C-positive dialysis patient from a county-level hospital to a tertiary hospital's dialysis center.

[0084] Step 1: Initiating a referral Doctors at dialysis centers in county-level hospitals submit referral applications on the medical consortium referral collaboration platform, filling in the following information: patient's name, ID number, medical card number, dialysis history (dialysis duration of 3 years, HD 3 times per week), recent infectious disease test record (infectious disease test result 3 months ago, hepatitis C positive), and reason for referral.

[0085] Step 2: Pre-detection trigger The administrators and doctors of the dialysis center at a top-tier hospital received a referral request notification via a mini-program on the platform. According to the Standard Operating Procedure (SOP) for infectious disease management, more than three months had passed since the last infectious disease test, requiring a retest. The doctor issued an infectious disease testing order in the hospital's outpatient system based on the patient's medical card number and simultaneously sent a notification to the patient's mobile phone via the platform: "Your infectious disease testing order has been issued. Please go to the outpatient department of a certain top-tier hospital before [date] to complete the blood test." After the test results were completed (hepatitis C RNA positive, confirming the patient's hepatitis C infection), the doctor entered the referred patient's dialysis treatment order based on the infectious disease result: HD 3 times / week, HDF 2 times / month, Hp 2 times / month. The platform automatically matched the bed allocation engine rules according to the order, locking the patient's infectious disease zone identifier as the hepatitis C isolation zone and matching the corresponding treatment bed and treatment time within that zone.

[0086] Step 3: Arrange beds in advance and notify others Based on the availability of beds in the hepatitis C isolation ward of the tertiary hospital's dialysis center, the bed allocation engine automatically pre-assigns a bed for the patient (e.g., bed 3 in the hepatitis C isolation ward, morning shift) and generates a pre-assignment record (status: pre-assigned). The pre-assignment result is simultaneously pushed to the patient's mini-program via the platform: You have successfully pre-assigned a bed for the morning shift on [date], in bed 3 of the hepatitis C isolation ward. Please arrive at the tertiary hospital's dialysis center before 7:30 AM on that day for dialysis.

[0087] Step 4: Confirm arrival and on-machine operation The patient arrives at the dialysis center nursing station of the tertiary hospital at the notified time. After the nurse verifies the patient's identity on the platform, she clicks to confirm arrival, and the system switches the bed status from pre-assigned to locked. The patient does not need to register on-site or wait for infectious disease test results. After completing the pre-dialysis preparation, the patient can directly start treatment, with the entire waiting time not exceeding 15 minutes. Example 3

[0088] This example uses a dialysis patient with a history of hypertension and diabetes as an example. The patient's vascular access is an autogenous arteriovenous fistula (AVF), which details the specific implementation of the home monitoring and early warning linkage subsystem.

[0089] Step 1: Real-time collection and uploading of home data Patients measure their blood pressure and weight twice a day, morning and evening, and their fasting blood sugar before and after meals. They also assess their AVF function twice a day, morning and evening. The data is synchronized to the mini-program via voice and then uploaded to the hospital server. The system backend receives and stores the data in real time.

[0090] The system sets personalized monitoring and early warning thresholds based on the patient's dry weight (60kg) and baseline blood pressure (systolic pressure 120-140mmHg): weight fluctuation exceeding dry weight ±2kg triggers an early warning, systolic blood pressure <90mmHg or >180mmHg triggers an early warning, and disappearance of arteriovenous fistula thrill signal triggers an emergency early warning.

[0091] Step Two: Anomaly Warning and Classification Processing Scenario 1 (Critical Value Warning): One evening before dinner, the patient underwent a routine AVF function test. The AVF thrill signal was weak, unlike its usual strong and powerful nature. Blood pressure was measured at 86 / 51 mmHg. The monitoring data was immediately uploaded. Upon receiving the warning on the doctor's mobile app, it was determined to be a critical value indicating the disappearance of the AVF thrill. A red warning message (pop-up window + sound alert) was immediately sent to the dialysis center management terminal. Simultaneously, an emergency notification was pushed to the patient's mobile app: "Abnormal AVF thrill detected. Please seek medical attention immediately or contact the dialysis center." Upon receiving the warning, the on-duty nurse at the nursing station immediately called the patient to confirm the situation and coordinated with the emergency department or arranged for emergency extra drainage treatment that day to address the disappearance of the AVF thrill. Through timely intervention, the patient's AVF function returned to normal.

[0092] Scenario 2: Leave Request Triggers Bed Allocation On a Tuesday morning, a patient clicks to apply for leave in the patient-side mini-program, selects the treatment time slot for that day, and fills in a note stating that there is a family matter, rescheduling dialysis from Wednesday morning to Thursday. Upon receiving the leave application, the system automatically marks the patient's pre-assigned bed for that day as released, and simultaneously sends a patient leave-bed release notification to the nursing station. This bed becomes available for other waiting patients, and the system simultaneously removes the released consumables from the pre-discharge list.

[0093] Step 3: Feedback on the linkage results Once a bed is released, it remains vacant, and nurses will no longer prepare consumables for that bed.

[0094] All early warning processing results and bed scheduling linkage records are automatically archived to the patient management file, which doctors can access in the system, forming a complete data closed loop for the entire cycle of home monitoring, early warning linkage, bed scheduling adjustment and treatment execution.

[0095] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A collaborative management system for resources in a hemodialysis medical consortium based on intelligent bed scheduling decision-making, characterized in that, include: The intelligent decision-making bed scheduling management subsystem is used to acquire the medical order information of dialysis patients, construct a multi-constraint rule matrix containing rigid constraints and flexible optimization conditions, and perform joint matching of beds and dialysis equipment through the bed scheduling algorithm engine based on the multi-constraint rule matrix and the medical order information to generate an initial bed scheduling plan. It also listens for dynamic events, and when the dynamic event is triggered, it incrementally rearranges the initial bed scheduling plan based on the multi-constraint rule matrix to generate and output the target bed scheduling plan. The medical consortium referral collaboration platform is used to receive referral application information sent by primary healthcare institutions, trigger the pre-test process to obtain infectious disease test results, and then write the bed allocation algorithm engine triggered by the doctor's order to pre-allocate beds according to available beds. After receiving the hospital's confirmation instruction, the pre-allocated beds are switched to a locked state. The home monitoring and early warning linkage subsystem is used to collect home monitoring data uploaded by patient terminals, compare the home monitoring data with preset thresholds to determine whether an abnormal early warning is triggered, and when a critical value early warning is triggered, the bed scheduling algorithm engine is activated to arrange an expedited bed. When an intention to request leave is detected, the bed scheduling algorithm engine is activated to release a bed.

2. The hemodialysis medical consortium resource collaborative management system based on intelligent bed scheduling decision-making as described in claim 1, characterized in that, Also includes: The system includes a management terminal, a nurse terminal, a doctor terminal, and a patient terminal, all of which are communicatively connected to the intelligent bed scheduling management subsystem. The management terminal is used for system configuration, importing and exporting medical orders, setting bed scheduling rules, verifying and adjusting bed scheduling, managing patients referred by medical consortia, managing home monitoring data, and viewing data statistics; The nurse terminal is used to view patient bed assignment information, mark patients on leave, make temporary adjustments to bed assignments for the day, and mark blood collection information; The doctor's terminal is used to enter medical orders, view patient bed arrangement information, view home monitoring data and abnormal alerts, approve patient leave applications and medical consortium referral applications, and view infectious disease test results. The patient terminal is used to view personal bed arrangements, upload home health data, submit changes in treatment needs or leave applications, and view home health education content.

3. A method for collaborative management of resources in a hemodialysis medical consortium based on intelligent bed scheduling decision-making, characterized in that, include: Obtain medical orders from dialysis patients; Construct a multi-constraint rule matrix that includes both rigid constraints and flexible optimization conditions; Based on the multi-constraint rule matrix and the medical order information, the bed allocation algorithm engine performs joint matching of beds and dialysis equipment to generate an initial bed allocation plan. Listen for dynamic events. When a treatment need is triggered by a dynamic event, incrementally rearrange the initial bed scheduling plan based on the multi-constraint rule matrix to generate and output the target bed scheduling plan.

4. The method for collaborative management of resources in a hemodialysis medical consortium based on intelligent bed scheduling decision-making as described in claim 3, characterized in that, The rigid constraints include infectious disease zoning and isolation constraints, treatment mode and equipment function compatibility constraints, and time period capacity constraints. Based on the multi-constraint rule matrix and the medical order information, the bed allocation algorithm engine performs joint matching of beds and dialysis equipment to generate an initial bed allocation plan, including: Based on the infectious disease identifier in the medical order information, the corresponding isolation zone is matched; Verify whether the number of patients admitted to each isolation zone during the target treatment period exceeds the corresponding bed capacity limit; Based on the dialysis treatment parameters in the medical order information, dialysis devices that support the corresponding dialysis treatment mode and are in normal condition are selected within the matched isolation zone, and the physical beds corresponding to the selected dialysis devices are allocated to the dialysis patients.

5. The method for collaborative management of resources in a hemodialysis medical consortium based on intelligent bed scheduling decision-making as described in claim 3, characterized in that, The flexible optimization conditions include scheduling strategy optimization conditions and treatment period and bed continuity preference conditions; The step of generating an initial bed allocation plan by performing joint matching of beds and dialysis equipment through a bed allocation algorithm engine based on the multi-constraint rule matrix and the medical order information also includes: When the medical order information includes a hemodialysis filtration or hemoperfusion treatment mode, the treatment mode is fixed on a fixed date every week according to a periodic scheduling strategy, or the number of days between each treatment of the treatment mode is evenly distributed according to a scheduling strategy that evenly distributes the number of treatments in the month. Provided that the rigid constraints are met, dialysis patients are preferentially assigned to their historically preferred beds and treatment times.

6. The method for collaborative management of resources in a hemodialysis medical consortium based on intelligent bed scheduling decision-making as described in claim 3, characterized in that, The dynamic events include medical order change events, patient leave events, equipment failure events, and patient status change events; When the dynamic event treatment demand is triggered, the initial bed scheduling plan is incrementally rearranged based on the multi-constraint rule matrix to generate and output a target bed scheduling plan, including: When the medical order change event is detected, the original bed allocation plan for the corresponding patient is withdrawn, and the bed allocation is re-executed according to the changed medical order information; When the patient's leave event is detected, the target bed for the target time period allocated to the corresponding patient during the leave period is marked as released. When the equipment failure event is detected, the corresponding dialysis patient originally assigned to the faulty equipment is transferred to an available bed that meets the rigid constraint conditions; When the patient status change event is detected, the subsequent bed scheduling plan for the corresponding patient is cleared.

7. The method for collaborative management of resources in a hemodialysis medical consortium based on intelligent bed scheduling decision-making as described in claim 3, characterized in that, It also includes the medical consortium referral and coordination process: Receive referral requests from primary healthcare institutions; After triggering the pre-detection process and obtaining the infectious disease test results, the results are written into the bed scheduling algorithm engine through medical orders; Pre-allocation is performed based on available beds, and the pre-allocation results are pushed to the patient's terminal; Upon receiving confirmation from the hospital, the pre-assigned bed will be switched to a locked state.

8. The method for collaborative management of resources in a hemodialysis medical consortium based on intelligent bed scheduling decision-making as described in claim 7, characterized in that, The pre-detection process, after obtaining the infectious disease test results, involves writing medical orders into the bed scheduling algorithm engine, including: If test results are mutually recognized within the region, the partition identifier of recent infectious disease test records can be directly extracted. After the doctor confirms, the medical order is written into the bed scheduling algorithm engine. If test results are not mutually recognized within the region, the medical order is connected to the outpatient system and a test order is generated. After receiving the qualified infectious disease test results from the retest, the medical order is written into the bed scheduling algorithm engine.

9. The method for collaborative management of resources in a hemodialysis medical consortium based on intelligent bed scheduling decision-making as described in claim 3, characterized in that, It also includes the home monitoring and bed arrangement linkage process: Collect home monitoring data uploaded by patients' terminals; The home monitoring data is compared with personalized thresholds to determine whether an abnormal warning has been triggered. When a critical value warning is triggered, the patient's needs will be coordinated with the bed scheduling algorithm engine to arrange an expedited bed; When a leave request is detected, the bed allocation algorithm engine is activated to release a bed.

10. The method for collaborative management of resources in a hemodialysis medical consortium based on intelligent bed scheduling decision-making as described in claim 3, characterized in that, It also includes multi-terminal role collaboration and statistical backtracking processes: Based on the role permissions of management terminals, nurse terminals, doctor terminals, and patient terminals, bed allocation information is pushed to specific users. Record bed utilization rate, equipment usage data, patient change data and consumable usage data, generate multi-dimensional statistical backtracking charts, and trigger dialysis indicator monitoring reminders according to preset cycles, and push the list of patients who should have blood drawn for monitoring to each terminal.

Images