Medical equipment health management system

Abstract

The invention belongs to the technical field of medical equipment management, and discloses a medical equipment health management system. Comprising a multi-dimensional health data acquisition module, an intelligent diagnosis engine module, a predictive maintenance scheduling module, a cross-device health collaboration module, a full-life-cycle health archive module, a user interaction and early warning module, a background management center and an emergency health guarantee module. Through the digital twin model and the LSTM-GRU hybrid deep learning algorithm, the equipment state can be accurately mapped in real time, the potential fault can be predicted in advance, the fault root cause can be quickly positioned in combination with the knowledge graph, and the medical service interruption caused by the sudden equipment fault can be effectively avoided; the predictive maintenance scheduling module can dynamically generate a maintenance plan, and in combination with the maintenance resource dynamic optimization unit, an optimal path can be intelligently planned according to an engineer state, spare part inventory and the like, so that the maintenance efficiency is improved, the maintenance cost is reduced, meanwhile, the equipment is ensured to serve clinic in a good state, and actual application and operation are facilitated.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of medical equipment management, in particular to a medical equipment health management system. BACKGROUND

[0002] In the modern medical system, medical equipment is an important basis for carrying out clinical diagnosis, treatment and other work, and its performance stability is directly related to the quality of medical services and patient safety.

[0003] However, in the actual application process of the existing equipment, most of them only focus on part of the parameters of the equipment operation, lack comprehensive collection of multi-dimensional data (such as core component working conditions, environmental influences, etc.), and are difficult to accurately reflect the real health status of the equipment, and are mostly dependent on manual inspection or troubleshooting after equipment failure occurs, which cannot predict potential failures in advance, often leading to forced interruption of medical work when a fault occurs, seriously affecting the diagnosis and treatment efficiency, and even may cause medical risks. Maintenance plan is mostly regular and fixed, without dynamic adjustment according to the actual health status of the equipment and clinical priority, which is easy to cause waste of maintenance resources, or when the equipment needs maintenance, resources cannot be allocated in time, and there is a lack of effective health correlation analysis between multiple associated medical equipment in the same clinical scene. When a single device is abnormal, it is difficult to quickly assess the impact on associated devices, and data of the same type of equipment across hospital areas is difficult to share and train, the diagnosis model has poor generalization ability, and the whole process health data of the equipment from factory to scrap is not effectively integrated and utilized, which cannot provide comprehensive basis for the update iteration and cost evaluation of the equipment. SUMMARY

[0004] One of the purposes of the present application is to provide a medical equipment health management system.

[0005] To achieve the above purpose, the technical scheme adopted by the present application is as follows: a medical equipment health management system, comprising a multi-dimensional health data acquisition module, an intelligent diagnosis engine module, a predictive maintenance scheduling module, a cross-equipment health collaboration module, a whole life cycle health record module, a user interaction and early warning module, a background management hub and an emergency health guarantee module;

[0006] The multi-dimensional health data acquisition module acquires real-time operation state data, core component working condition data, environmental influence data and historical maintenance record data of the medical equipment, wherein the operation state data includes real-time power, operating temperature, data transmission delay and key operation response speed of the equipment, the core component working condition data includes component vibration frequency, wear degree parameter and voltage stability, and the environmental influence data includes temperature and humidity, dust concentration and electromagnetic interference intensity of the environment where the equipment is located;

[0007] The intelligent diagnosis engine module is in communication connection with the multi-dimensional health data acquisition module, constructs a device virtual mirror based on a digital twin model, maps the running state of a physical entity of the device in real time, analyzes the collected data by using an LSTM-GRU hybrid deep learning algorithm, realizes prediction of potential faults of the device and determination of fault types and fault levels, and locates a fault root cause in combination with a device component correlation knowledge graph;

[0008] The predictive maintenance scheduling module is in communication connection with the intelligent diagnosis engine module and the background management hub respectively, dynamically generates a device maintenance plan according to the fault prediction result and the fault level output by the intelligent diagnosis engine module, intelligently allocates maintenance resources to schedule maintenance tasks based on a real-time state of the device clinical priority and the maintenance resources.

[0009] The cross-device health cooperation module is in communication connection with the multi-dimensional health data acquisition module and the intelligent diagnosis engine module, establishes a health correlation graph of multiple associated medical devices in the same clinical scene, analyzes the influence of an abnormal state of a single device on the health of associated devices, and realizes cooperative training of health data of the same type of devices across hospital areas by using a federated learning technology, thereby improving the generalization ability of a diagnosis model.

[0010] The whole-life-cycle health record module is in communication connection with the multi-dimensional health data acquisition module, the intelligent diagnosis engine module, and the predictive maintenance scheduling module, stores whole-process health data of a device from factory shipment, installation and debugging, clinical use, maintenance and repair to scrapping, generates a device health trend curve, calculates a device maintenance cost coefficient and a residual value evaluation value based on health data, and provides a decision basis for device updating and iteration.

[0011] The user interaction and early warning module is in communication connection with the intelligent diagnosis engine module and the predictive maintenance scheduling module respectively, outputs graded early warning information according to the fault level, first-level early warning: possible shutdown within 2 hours, second-level early warning: performance decline within 24 hours, third-level early warning: need to be concerned within 72 hours, and fourth-level early warning: regular health fluctuation, and pushes early warning information, a maintenance plan and a device health report to a medical operation end, a maintenance engineer end and a hospital management end.

[0012] The background management hub is in communication connection with all the modules, aggregates, stores and manages data of the modules, and supports an administrator to configure diagnosis model parameters, early warning thresholds and maintenance priority rules.

[0013] Preferably, the multi-dimensional health data acquisition module includes an edge computing node arranged in a medical device, the edge computing node performs real-time preprocessing on collected raw data, including data denoising, outlier rejection and data compression, and then transmits the preprocessed data to the intelligent diagnosis engine module, thereby reducing data transmission bandwidth occupation and background computing pressure.

[0014] Preferably, the fault root cause localization process of the intelligent diagnostic engine module is as follows: based on the knowledge graph of equipment components association, the determined fault type is taken as the starting node, the core components, operating parameters, environmental factors and historical fault case nodes related to the fault in the graph are traversed, the association weight between each node and the fault type node is calculated, the nodes with association weight ≥ 0.8 are selected as candidate root causes, and then the authenticity of the candidate root causes is verified by combining the real-time collected core component operating condition data and environmental data, and finally the fault root cause is determined.

[0015] Preferably, the predictive maintenance scheduling module further includes a maintenance resource dynamic optimization unit. The maintenance resource dynamic optimization unit obtains the current location, skill certification information and current task progress of the maintenance engineer in real time, and combines the real-time location of spare parts inventory and traffic congestion data of delivery routes. It uses the improved Dijkstra algorithm to calculate the optimal scheduling path from maintenance resources to the target equipment to ensure that the maintenance task is completed within the warning threshold.

[0016] Preferably, the cross-device health collaboration module further includes an associated device abnormal linkage response unit. When a device triggers a level 2 or higher warning, the associated device abnormal linkage response unit automatically retrieves the associated devices in the health association map, sends a temporary health monitoring frequency increase instruction to the associated devices, and simultaneously synchronizes the real-time health data of the associated devices to the intelligent diagnostic engine module to analyze whether there is a chain of abnormal risks. If there is a chain of risks, the associated devices are triggered with graded warnings.

[0017] Preferably, the full life cycle health record module also includes an equipment health cost analysis unit. The equipment health cost analysis unit calculates the maintenance costs, downtime losses and energy consumption costs during the entire life cycle of the equipment, generates a unit health duration cost coefficient, compares it with the industry average of similar equipment, and outputs an equipment health economic assessment report.

[0018] Preferably, the user interaction and early warning module also includes an early warning response closed-loop unit. When an early warning information is pushed, the early warning response closed-loop unit tracks the confirmation status of medical staff in real time, whether they check the early warning, the task reception status of maintenance engineers, and the maintenance completion status. If no corresponding response is received within 30 minutes for a Level 1 early warning, 2 hours for a Level 2 early warning, and 6 hours for a Level 3 early warning, the early warning level is automatically upgraded and pushed to a higher-level management terminal. At the same time, an early warning response delay report is generated to optimize the hospital equipment management process.

[0019] Preferably, the back-end management center further includes a diagnostic model self-iteration unit. The diagnostic model self-iteration unit periodically compares the actual fault data of the equipment with the prediction results of the intelligent diagnostic engine module, calculates the model prediction accuracy and root cause localization accuracy. When the accuracy is lower than the preset threshold (prediction accuracy < 90%) and the root cause localization accuracy < 85%), the actual fault data is automatically added to the model training set, and the parameters of the LSTM-GRU hybrid deep learning model are updated using an incremental learning algorithm to achieve model self-optimization.

[0020] Preferably, the emergency health protection module is communicatively connected to the back-end management center. When a public health emergency occurs, the emergency health protection module automatically retrieves medical equipment related to emergency treatment within the hospital, quickly generates an emergency equipment health status report, prioritizes marking equipment with good health status and assigning it to the emergency treatment area, and triggers mandatory health monitoring for equipment with excessive operating load to prevent equipment failure due to overload.

[0021] Preferably, the emergency health protection module also includes a backup equipment linkage scheduling unit. When the equipment in the emergency treatment area triggers a level one warning and there are no immediate maintenance conditions, the backup equipment linkage scheduling unit automatically searches for the health status of similar backup equipment in the hospital, generates a backup equipment allocation route, pushes the equipment handover list to medical staff, and updates the emergency use records of the equipment in the full life cycle health record.

[0022] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0023] (1) This invention uses a digital twin model and a hybrid deep learning algorithm of LSTM-GRU to accurately map the status of equipment in real time, predict potential faults in advance, and quickly locate the root cause of the fault by combining knowledge graphs, effectively avoiding medical service interruptions caused by sudden equipment failures; the predictive maintenance scheduling module can dynamically generate maintenance plans, and combined with the maintenance resource dynamic optimization unit, it can intelligently plan the optimal path according to the engineer status, spare parts inventory, etc., improve maintenance efficiency, reduce maintenance costs, and at the same time ensure that the equipment serves the clinic in good condition.

[0024] (2) This invention, through the cross-device health collaboration module, establishes a health association map and uses federated learning technology. It can not only analyze the impact of a single device's abnormality on related devices, but also collaboratively train data of similar devices across different hospital areas, improve the applicability of the diagnostic model in different scenarios, and enhance the overall synergy of medical equipment management. The full life cycle health record module fully records data of each stage of the device and generates health trend curves, maintenance cost coefficients, etc., providing a scientific basis for hospital equipment updates and cost control, and helping hospitals optimize equipment asset allocation.

[0025] (3) The present invention enables personnel at different levels to know the status of equipment in a timely manner through user interaction and the hierarchical early warning module. The early warning response closed-loop unit ensures that the early warning is processed in a timely manner, avoiding more serious problems caused by response delays, and improving the standardization and timeliness of hospital equipment management. The emergency health protection module can quickly allocate health equipment during public health emergencies to ensure emergency diagnosis and treatment needs. The diagnostic model self-iteration unit of the back-end management center can continuously optimize the model, so that the system's diagnostic and prediction capabilities can be continuously improved and adapted to changes after long-term operation of equipment. Attached Figure Description

[0026] Figure 1 This is a schematic diagram of the overall process structure of the present invention. Detailed Implementation

[0027] The present invention will now be further described in conjunction with specific embodiments. It should be noted that, without conflict, the various embodiments or technical features described below can be arbitrarily combined to form new embodiments.

[0028] In the description of this invention, it should be noted that directional terms such as "center," "lateral," "longitudinal," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise" indicate the orientation and positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. They should not be construed as limiting the specific protection scope of this invention.

[0029] It should be noted that the terms "first" and "second" in the specification and claims of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0030] One preferred embodiment of the present invention, such as Figure 1 As shown, a medical device health management system includes a multi-dimensional health data acquisition module, an intelligent diagnostic engine module, a predictive maintenance scheduling module, a cross-device health collaboration module, a full life cycle health record module, a user interaction and early warning module, a back-end management center, and an emergency health protection module.

[0031] The multi-dimensional health data acquisition module collects real-time data on the operating status of medical equipment, the operating conditions of core components, environmental impact data, and historical maintenance records. The operating status data includes real-time power, operating temperature, data transmission latency, and response speed of key operations. The operating conditions of core components include component vibration frequency, wear parameters, and voltage stability. The environmental impact data includes temperature and humidity, dust concentration, and electromagnetic interference intensity of the environment in which the equipment is located.

[0032] The intelligent diagnostic engine module communicates with the multi-dimensional health data acquisition module, constructs a virtual image of the equipment based on the digital twin model, maps the operating status of the physical entity of the equipment in real time, and uses the LSTM-GRU hybrid deep learning algorithm to analyze the collected data, realize the prediction of potential equipment faults and the determination of fault type and fault level. At the same time, it combines the knowledge graph of equipment component association to complete the location of the root cause of the fault.

[0033] The predictive maintenance scheduling module is connected to the intelligent diagnostic engine module and the back-end management center. Based on the fault prediction results and fault levels output by the intelligent diagnostic engine module, it dynamically generates equipment maintenance plans and intelligently allocates maintenance resources and schedules maintenance tasks based on the clinical priority of the equipment and the real-time status of maintenance resources.

[0034] The cross-device health collaboration module communicates with the multi-dimensional health data acquisition module and the intelligent diagnostic engine module to establish a health association map of multiple related medical devices in the same clinical scenario, analyze the impact of abnormal status of a single device on the health of related devices, and at the same time use federated learning technology to achieve collaborative training of health data of similar devices across hospitals, thereby improving the generalization ability of the diagnostic model.

[0035] The full life cycle health record module communicates with the multi-dimensional health data acquisition module, intelligent diagnostic engine module, and predictive maintenance scheduling module. It stores the health data of the device throughout the entire process from manufacturing, installation and commissioning, clinical use, maintenance and repair to scrapping, generates equipment health trend curves, calculates equipment maintenance cost coefficients and residual value assessments based on health data, and provides decision-making basis for equipment updates and iterations.

[0036] The user interaction and early warning module communicates with the intelligent diagnostic engine module and the predictive maintenance scheduling module, respectively. It outputs graded early warning information according to the fault level: Level 1 warning: possible downtime within 2 hours; Level 2 warning: performance degradation within 24 hours; Level 3 warning: attention required within 72 hours; Level 4 warning: routine health fluctuations. The early warning information, maintenance plan and equipment health report are pushed to the medical staff operation terminal, maintenance engineer terminal and hospital management terminal.

[0037] The backend management center communicates with all the above modules to summarize, store, and manage the data of each module, and also supports administrators in configuring diagnostic model parameters, early warning thresholds, and maintenance priority rules.

[0038] The multi-dimensional health data acquisition module includes edge computing nodes deployed inside medical devices. The edge computing nodes perform real-time preprocessing on the collected raw data, including data noise reduction, outlier removal, and data compression, and then transmit the preprocessed data to the intelligent diagnostic engine module, reducing data transmission bandwidth usage and backend computing pressure.

[0039] The fault root cause localization process of the intelligent diagnostic engine module is as follows: Based on the knowledge graph of equipment component association, the determined fault type is taken as the starting node, and the core components, operating parameters, environmental factors and historical fault case nodes related to the fault in the graph are traversed. The association weight between each node and the fault type node is calculated, and nodes with an association weight ≥ 0.8 are selected as candidate root causes. Then, the authenticity of the candidate root causes is verified by combining the real-time collected core component operating condition data and environmental data, and finally the fault root cause is determined.

[0040] The predictive maintenance scheduling module also includes a maintenance resource dynamic optimization unit. This unit obtains the current location, skill certification information, and current task progress of maintenance engineers in real time. Combined with the real-time location of spare parts inventory and traffic congestion data of delivery routes, it uses the improved Dijkstra algorithm to calculate the optimal scheduling path from maintenance resources to the target equipment, ensuring that maintenance tasks are completed within the warning threshold.

[0041] The cross-device health collaboration module also includes an associated device anomaly linkage response unit. When a device triggers a level 2 or higher warning, the associated device anomaly linkage response unit automatically retrieves the associated devices in the health association map, sends a temporary health monitoring frequency increase instruction to the associated devices, and synchronizes the real-time health data of the associated devices to the intelligent diagnostic engine module to analyze whether there is a chain of anomaly risks. If there is a chain of risks, it triggers a graded warning for the associated devices.

[0042] The full life cycle health record module also includes an equipment health cost analysis unit. This unit calculates maintenance costs, downtime losses, and energy costs throughout the equipment's life cycle, generates a cost coefficient per unit of health duration, compares it with the industry average for similar equipment, and outputs an equipment health economic assessment report.

[0043] The user interaction and early warning module also includes an early warning response closed-loop unit. When an early warning information is pushed, the early warning response closed-loop unit tracks the confirmation status of medical staff in real time, whether they have checked the early warning, the task reception status of maintenance engineers, and the maintenance completion status. If no corresponding response is received within 30 minutes for a Level 1 early warning, 2 hours for a Level 2 early warning, and 6 hours for a Level 3 early warning, the early warning level will be automatically upgraded and pushed to a higher-level management terminal. At the same time, an early warning response delay report will be generated to optimize the hospital equipment management process.

[0044] The back-end management center also includes a diagnostic model self-iteration unit. The diagnostic model self-iteration unit regularly compares the actual fault data of the equipment with the prediction results of the intelligent diagnostic engine module, calculates the model prediction accuracy and root cause location accuracy. When the accuracy is lower than the preset threshold (prediction accuracy < 90%) and the root cause location accuracy < 85%), the actual fault data is automatically added to the model training set, and the parameters of the LSTM-GRU hybrid deep learning model are updated using an incremental learning algorithm to achieve model self-optimization.

[0045] The emergency health protection module is connected to the back-end management center. When a public health emergency occurs, the emergency health protection module automatically retrieves medical equipment related to emergency treatment in the hospital, quickly generates a health status report of the emergency equipment, prioritizes marking equipment with good health status and assigning it to the emergency treatment area, and triggers mandatory health monitoring for equipment with excessive operating load to prevent equipment failure due to overload.

[0046] The emergency health protection module also includes a backup equipment linkage dispatch unit. When the equipment in the emergency treatment area triggers a level one warning and there are no immediate maintenance conditions, the backup equipment linkage dispatch unit automatically searches for the health status of similar backup equipment in the hospital, generates a backup equipment allocation route, pushes the equipment handover list to medical staff, and updates the emergency use records of the equipment in the full life cycle health record.

[0047] Working principle:

[0048] In use, the management system first collects multi-dimensional data on the medical equipment's operating status, core component conditions, environmental impact, and historical maintenance records through a multi-dimensional health data acquisition module using edge computing nodes, and preprocesses the raw data. Next, the intelligent diagnostic engine module constructs a virtual image of the equipment based on a digital twin model, uses an LSTM-GRU hybrid deep learning algorithm to analyze the preprocessed data, and combines it with a component association knowledge graph to achieve fault prediction, type and level determination, and root cause localization. The results are distributed to the user early warning module, maintenance scheduling module, and cross-device collaboration module: the user early warning module outputs tiered early warnings and pushes relevant information, tracking responses to form a closed loop; the maintenance scheduling module dynamically generates maintenance plans based on fault results, intelligently allocates resources, and its maintenance resource dynamic optimization unit also calculates the optimal scheduling path; the cross-device collaboration module establishes associations. The graph analysis of equipment anomalies enhances the model's generalization ability through federated learning and collaborative training. The associated equipment anomaly response unit increases the monitoring frequency of associated equipment and analyzes cascading risks when equipment triggers a level 2 or higher warning. Data from these modules is stored in the full lifecycle health record module, generating health trend curves and providing a basis for equipment updates. The backend management center aggregates and stores data from each module, manages access permissions, and supports parameter configuration. Its diagnostic model self-iteration unit periodically compares actual fault data with prediction results and updates model parameters. In the event of an emergency, the emergency health protection module automatically retrieves relevant emergency equipment, generates a health status report, and allocates equipment, triggering mandatory monitoring for overloaded equipment. The backup equipment linkage scheduling unit schedules similar backup equipment and updates its records when emergency equipment triggers a level 1 warning and there are no immediate maintenance conditions.

[0049] The basic principles, main features, and advantages of this invention have been described above. Those skilled in the art should understand that this invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made without departing from the spirit and scope of the invention, and all such changes and modifications fall within the scope of the invention as claimed. The scope of protection claimed by this invention is defined by the appended claims and their equivalents.

Claims

1. A medical device health management system, comprising: The system comprises a multi-dimensional health data acquisition module, an intelligent diagnosis engine module, a predictive maintenance scheduling module, a cross-device health collaboration module, a full life cycle health record module, a user interaction and early warning module, a background management hub, and an emergency health guarantee module. The multi-dimensional health data acquisition module acquires real-time operation state data, core component working condition data, environmental influence data, and historical maintenance record data of the medical device, wherein the operation state data includes real-time power, operating temperature, data transmission delay, and key operation response speed, the core component working condition data includes component vibration frequency, wear degree parameters, and voltage stability, and the environmental influence data includes temperature and humidity, dust concentration, and electromagnetic interference intensity of the environment where the device is located. The intelligent diagnosis engine module is in communication connection with the multi-dimensional health data acquisition module, constructs a device virtual mirror based on a digital twin model, maps the operation state of the device physical entity in real time, analyzes the collected data by using an LSTM-GRU hybrid deep learning algorithm, realizes the prediction of potential faults and the determination of fault types and fault levels, and locates the fault root cause in combination with a device component correlation knowledge graph. The predictive maintenance scheduling module is in communication connection with the intelligent diagnosis engine module and the background management hub, dynamically generates a device maintenance plan according to the fault prediction result and fault level output by the intelligent diagnosis engine module, intelligently allocates maintenance resources to schedule maintenance tasks based on the real-time state of the device clinical priority and maintenance resources. The cross-device health collaboration module is in communication connection with the multi-dimensional health data acquisition module and the intelligent diagnosis engine module, establishes a health correlation graph of multiple associated medical devices in the same clinical scene, analyzes the influence of the abnormal state of a single device on the health of associated devices, and realizes the collaborative training of health data of the same type of devices across hospital areas by using a federated learning technology, thereby improving the generalization ability of the diagnosis model. The full life cycle health record module is in communication connection with the multi-dimensional health data acquisition module, the intelligent diagnosis engine module, and the predictive maintenance scheduling module, stores the full-process health data of the device from factory shipment, installation and debugging, clinical use, maintenance and repair to scrapping, generates a device health trend curve, calculates the device maintenance cost coefficient and residual value evaluation value based on the health data, and provides a decision basis for device update and iteration. The user interaction and early warning module is in communication connection with the intelligent diagnosis engine module and the predictive maintenance scheduling module, outputs graded early warning information according to the fault level, the first-level early warning is that the device may stop within 2 hours, the second-level early warning is that the performance of the device will decrease within 24 hours, the third-level early warning is that the device needs to be paid attention to within 72 hours, and the fourth-level early warning is that the device has a regular health fluctuation, and the early warning information, maintenance plan, and device health report are pushed to the medical operation end, the maintenance engineer end, and the hospital management end. The background management hub is in communication connection with all the modules, aggregates, stores, and manages the data of the modules, and supports the administrator to configure the diagnosis model parameters, early warning threshold values, and maintenance priority rules.

2. The medical device health management system of claim 1, wherein: The multi-dimensional health data acquisition module includes an edge computing node arranged in the medical device, the edge computing node performs real-time preprocessing on the collected raw data, including data denoising, outlier rejection and data compression, and then transmits the preprocessed data to the intelligent diagnosis engine module, thereby reducing data transmission bandwidth occupation and background computing pressure.

3. The medical device health management system of claim 1, wherein: The fault root cause positioning process of the intelligent diagnosis engine module is as follows: based on the device component association knowledge graph, the determined fault type is taken as a starting node, the core components, operating parameters, environmental factors and historical fault case nodes related to the fault in the graph are traversed, the association weight of each node and the fault type node is calculated, the nodes with an association weight greater than or equal to 0.8 are selected as candidate root causes, and the real-time collected core component working condition data and environmental data are combined to verify the authenticity of the candidate root causes, and finally the fault root cause is determined.

4. The medical device health management system of claim 1, wherein: The predictive maintenance scheduling module further includes a maintenance resource dynamic optimization unit, which obtains the current position, skill certification information and current task progress of the maintenance engineer in real time, combines the real-time position of the spare parts inventory and the traffic congestion data of the distribution route, and calculates the optimal scheduling path of the maintenance resource to the target device by using the Dijkstra improved algorithm, thereby ensuring that the maintenance task is completed within the warning threshold.

5. The medical device health management system of claim 1, wherein: The cross-device health cooperation module further includes an associated device abnormal linkage response unit, when a device triggers a secondary or higher warning, the associated device abnormal linkage response unit automatically retrieves the associated devices of the device in the health association graph, sends a temporary health monitoring frequency improvement instruction to the associated devices, synchronizes the real-time health data of the associated devices to the intelligent diagnosis engine module, analyzes whether there is a chain abnormal risk, and if there is a chain risk, triggers a hierarchical warning of the associated devices.

6. The medical device health management system of claim 1, wherein: The full life cycle health record module further includes a device health cost analysis unit, which calculates the maintenance cost, fault downtime loss and energy consumption cost of the device in the full life cycle, generates a unit health time cost coefficient, compares it with the industry average of the same type of device, and outputs a device health economic evaluation report.

7. The medical device health management system of claim 1, wherein: The user interaction and warning module further includes a warning response closed loop unit, which tracks the confirmation status of medical staff after the warning information is pushed, whether to view the warning, the task receiving status of the maintenance engineer and the maintenance completion status, and if no corresponding response is received within a preset time of 30 minutes for a first warning, 2 hours for a second warning and 6 hours for a third warning, the warning level is automatically upgraded and pushed to a higher level of management terminal, and a warning response delay report is generated, thereby optimizing the hospital equipment management process.

8. The medical device health management system of claim 1, wherein: The background management hub further comprises a diagnostic model self-iteration unit, which periodically compares the actual fault data of the equipment with the prediction results of the intelligent diagnosis engine module, calculates the model prediction accuracy and root cause positioning accuracy, and when the accuracy is lower than the preset threshold prediction accuracy < 90%, root cause positioning accuracy < 85%, automatically adds the actual fault data to the model training set, updates the parameters of the LSTM-GRU mixed deep learning model using the incremental learning algorithm, and realizes self-optimization of the model.

9. The medical device health management system of claim 1, wherein: The emergency health guarantee module is in communication connection with the background management hub, and when a public health emergency occurs, the emergency health guarantee module automatically retrieves medical equipment related to emergency diagnosis and treatment in the hospital, quickly generates an emergency equipment health status report, preferentially marks the equipment with good health status for distribution to the emergency diagnosis and treatment area, and triggers forced health monitoring for equipment with excessively high operating load to avoid equipment failure due to overload.

10. The medical device health management system of claim 1, wherein: The emergency health guarantee module further comprises a standby equipment linkage scheduling unit, which automatically retrieves the health status of standby equipment of the same type in the hospital when the equipment in the emergency diagnosis and treatment area triggers a first-level early warning without immediate maintenance conditions, generates a standby equipment deployment route, pushes a device handover list to medical staff, and updates the emergency use records of the equipment in the full life cycle health file.

Images