Medical emergency processing method, system and equipment based on big data and medium

Abstract

The invention provides a medical emergency processing method, system and device based on big data and a medium, and belongs to the technical field of medical information.The method comprises the steps that a medical emergency event is responded, and structured event description data is generated; dynamically collecting and cleaning data, and constructing a data graph; performing illness condition criticality grading on the wounded, predicting illness condition development trend by combining a medical image analysis result of the wounded, and outputting an injury condition grading report; based on an improved VRPTW model and a genetic algorithm model, generating a scheduling scheme from ambulance initial assignment to path planning to medical resource pre-allocation; the resource execution state is tracked in real time, when the response time exceeds the set amplitude of the expected response time, the reinforcement learning model is started to dynamically adjust the scheduling scheme based on the data graph, and an effect evaluation report is generated. According to the invention, through multi-mode signal receiving, data integration, intelligent injury assessment, resource scheduling optimization and dynamic adjustment, rapid, accurate and efficient medical emergency response is realized.

Description

Technical Field

[0001] This application belongs to the field of medical information technology, specifically relating to a medical emergency response method, system, equipment, and medium based on big data. Background Technology

[0002] The efficiency and accuracy of medical emergency response are crucial for saving lives and reducing disability. Traditional medical emergency response methods often rely on human judgment and experience-based decision-making, making it difficult to achieve rapid and accurate resource allocation and treatment plan formulation in complex and ever-changing emergency scenarios. Especially when facing large-scale emergencies such as natural disasters, traffic accidents, and public health events, traditional methods often prove inadequate and fail to meet the needs for efficient and scientific emergency response.

[0003] Existing medical emergency response systems still have many shortcomings in data integration, analysis and decision-making, and resource allocation. On the one hand, real-time collection, cleaning, and correlation of multi-source data present technical challenges, making it difficult to guarantee data accuracy and timeliness. On the other hand, traditional decision-making models are often based on static rules or simple algorithms, which are difficult to adapt to complex and ever-changing emergency scenarios, resulting in low resource allocation efficiency and poor treatment outcomes. Furthermore, during emergency response, the lack of real-time monitoring and dynamic adjustment mechanisms for resource execution status makes it difficult to cope with the need for resource reallocation in sudden situations.

[0004] Therefore, developing a big data-based medical emergency response method and system to achieve real-time integration and analysis of multi-source data, intelligent injury classification and prediction, and optimized scheduling and dynamic adjustment of resources is of great significance for improving the efficiency and accuracy of medical emergency response. Summary of the Invention

[0005] In a first aspect, embodiments of this application provide a medical emergency response method based on big data, comprising the following steps: S1. Receive and respond to medical emergency event signals in real time through a multimodal input channel, classify and grade medical emergency events, and generate structured event description data for medical emergency events; S2. Dynamically collect historical case data, medical resource status data, environmental data, and vital sign assessment data of the injured, and clean and correlate them in real time through stream processing to construct a three-dimensional data map of medical emergency events, medical resources, and environment; S3. Use a gradient boosting tree model to classify the severity of the injuries, combine the medical image analysis results generated by the CNN model to predict the trend of the disease development, and output an injury classification report. S4. Based on the improved VRPTW model and genetic algorithm model, with emergency priority as the constraint and hospital load balancing as the goal, a scheduling scheme is generated from the initial assignment of ambulances to route planning and then to the pre-allocation of medical resources. The scheme is dynamically adjusted as the ambulance location data is updated, and optimization is performed using a three-dimensional data map. S5. Track the execution status of resources in real time, and when the response time exceeds the set range of the expected response time, start the reinforcement learning model to dynamically adjust the scheduling scheme based on the three-dimensional data map and generate an effect evaluation report.

[0006] Furthermore, the specific steps of step S1 are as follows: S11. Parse emergency call recordings through a voice recognition interface and extract key fields including the location of the incident, the number of casualties, and the description of the injuries; S12. Connect to roadside intelligent equipment to obtain video streams of the accident scene, and use the YOLOv5 model to identify the distribution density of the injured; S13. Match event types according to the preset rule base, and classify medical emergency events according to the number of casualties, whether there are critically injured people, and whether they involve the agreed type of injury; S14. Generate structured event codes that include geographic location coordinates, timestamps, and medical emergency event classifications.

[0007] Furthermore, the specific steps of step S2 are as follows: S21. Obtain medical resource status data from the hospital HIS system and emergency center dispatch system via HTTP / HTTPS interfaces; S22. Obtain meteorological warning data through the meteorological bureau's API interface, obtain real-time traffic flow through the map application interface, and use the meteorological warning data and real-time traffic flow as environmental data; S23. Obtain historical case data from a cross-institutional medical data center, wherein the historical case data includes resource scheduling schemes and treatment effect statistics for similar events; S24. Determine vital sign assessment data for the injured based on structured event description data; S25. Use the Apache Flink framework to perform real-time cleaning of multi-source data. Filter invalid characters using regular expressions, remove abnormal device data based on thresholds, and establish multi-dimensional relationships between medical emergency events, medical resources, and the environment through association rules to generate a three-dimensional data map containing spatial coordinates, time series, and resource attributes.

[0008] Furthermore, the medical resource status data includes real-time hospital bed occupancy, initial ambulance location, real-time ambulance location, on-duty status of medical staff, and professional skill tags.

[0009] Furthermore, the specific steps of step S3 are as follows: S31. Input the vital signs data, age and basic medical history of the injured into the gradient boosting model to determine the severity of the illness; S32. Use a CNN model to identify and classify lesions in the CT / MRI images of the injured, output the lesion probability value, and generate a severity rating report that includes a warning of the risk of deterioration; S33. Use a spatiotemporal graph neural network to fuse time-series data of vital signs and geospatial data to dynamically adjust the priority of medical resource matching corresponding to injury severity levels.

[0010] Furthermore, the specific steps of step S4 are as follows: S41. Based on injury severity reports, construct an emergency medical resource scheduling model with time windows, wherein: The golden treatment window for critically injured patients and the matching degree of specialists must meet the preset threshold as hard constraints. The optimization objective is to minimize the average response time and maximize the utilization rate of medical resources. S42. An improved Dijkstra algorithm is used to calculate the optimal path for a single wounded soldier. Real-time traffic data is introduced to dynamically update the edge weights, and the initial matching of wounded soldier and hospital is completed using the Hungarian algorithm. S43. Use the ambulance allocation scheme as the first gene segment and the path sequence as the second gene segment for chromosome encoding, and set the crossover probability and mutation probability of the genetic algorithm; S44. Monitor changes in medical resources and traffic data in the 3D data map in real time, and when the trigger threshold is met, retain the feasible region of the current optimal solution and start the local search algorithm for rapid adjustment; S45. Generate a set of executable dispatch instructions, including an ambulance dispatch list, dynamic navigation routes, and hospital resource readiness notifications.

[0011] Furthermore, the specific steps of step S5 are as follows: S51. Real-time collection of multi-dimensional monitoring indicators, including resource status data, casualty status data, and environmental update data; S52. Calculate the core deviation of each monitoring indicator, and initiate root cause analysis when the core deviation exceeds the set threshold; S53. Constructing dynamic policy optimization for Markov decision processes: Construct the state space S: S = {Injury severity rating, hospital load, road condition rating, remaining time window percentage}; Constructing action space A: A = {Reassign hospital, increase resources, change route, activate emergency protocol}; Construct the reward function:

[0012] in, The actual response time is represented by the actual response time, and the score represents the treatment outcome score. and For the pre-set weights, and ; Among them, the treatment outcome score is dynamically calculated based on the stability of the injured person's vital signs when they arrive at the hospital; S54. Use historical scheduling logs and the PPO algorithm to train the agent online, and update the policy network parameters once every set scheduling cycle; S55. Evaluate the effectiveness of emergency response processes based on timeliness indicators, medical quality indicators, and resource efficiency indicators.

[0013] Secondly, this application also provides a big data-based medical emergency response system, including an emergency event response module, which is used to receive and respond to medical emergency event signals in real time through a multimodal input channel, classify and grade medical emergency events, and generate structured event description data for medical emergency events. The multi-source data acquisition module is used to dynamically collect historical case data, medical resource status data, environmental data, and vital sign assessment data of the injured. It also performs real-time cleaning and correlation through stream processing to construct a three-dimensional data map of medical emergency events, medical resources, and the environment. The injury analysis module is used to classify the severity of the injuries using a gradient boosting tree model, combine the medical image analysis results generated by the CNN model to predict the development trend of the injury, and output an injury classification report. The resource allocation and scheduling module is used to generate a scheduling scheme based on the improved VRPTW model and genetic algorithm model, with emergency priority as the constraint and hospital load balancing as the goal, from the initial assignment of ambulances to route planning and then to the pre-allocation of medical resources. It is dynamically adjusted as the ambulance location data is updated and optimized using three-dimensional data maps. The scheduling result evaluation module is used to track the resource execution status in real time. When the response time exceeds the set range of the expected response time, the reinforcement learning model is activated to dynamically adjust the scheduling scheme based on the three-dimensional data map and generate an effect evaluation report.

[0014] Thirdly, embodiments of this application also provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the big data-based medical emergency response method described in the first aspect.

[0015] Fourthly, embodiments of this application also provide a storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the big data-based medical emergency response method described in the first aspect.

[0016] As can be seen from the above technical solutions, this application has the following advantages: The big data-based medical emergency response method, system, equipment, and media provided in this application receive medical emergency event signals in real time through a multimodal input channel, enabling rapid perception and response to emergencies, shortening the time from event occurrence to system activation, and saving valuable time for timely treatment of the injured. By dynamically collecting multi-source data and utilizing stream processing technology for real-time cleaning and correlation, a three-dimensional data map is constructed, solving the problems of data dispersion and difficulty in integration in traditional systems, and providing data support for subsequent decision-making. By combining gradient boosting tree models and CNN models, the severity of the injured's condition is graded and trends are predicted, improving the accuracy of injury assessment. Based on an improved VRPTW model and genetic algorithm model, a scheduling scheme is generated with the goals of emergency priority and hospital load balancing, and optimized using the three-dimensional data map, achieving rational allocation and efficient utilization of medical resources, avoiding resource waste and shortages. By tracking the resource execution status in real time, when situations such as response time exceeding expectations occur, a reinforcement learning model is activated to dynamically adjust the scheduling scheme, enhancing adaptability and flexibility in complex and ever-changing emergency scenarios, and improving the effectiveness and efficiency of emergency response. Attached Figure Description

[0017] To more clearly illustrate the technical solution of this application, the accompanying drawings used in the description will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart illustrating the big data-based medical emergency response method of the present invention.

[0019] Figure 2 This is a schematic diagram of the big data-based medical emergency response system of the present invention. Detailed Implementation

[0020] The various embodiments of this disclosure will be described more fully in the detailed steps of the big data-based medical emergency response method described below. This disclosure may have various embodiments, and adjustments and changes may be made therein. However, it should be understood that there is no intention to limit the various embodiments of this disclosure to the specific embodiments disclosed herein, but rather this disclosure should be understood to cover all adjustments, equivalents, and / or alternatives falling within the spirit and scope of the various embodiments of this disclosure.

[0021] For example, in medical emergency response, response speed and accuracy are crucial for saving lives and reducing disability. Traditional medical emergency response relies on human judgment and experience-based decision-making, which makes it difficult to quickly and accurately allocate resources and formulate treatment plans in complex and ever-changing emergency scenarios. The shortcomings of traditional methods become increasingly apparent when facing large-scale emergencies such as natural disasters, traffic accidents, and public health incidents, failing to meet the needs for efficient and scientific emergency response.

[0022] Currently, medical emergency response systems have significant shortcomings in data integration, analysis and decision-making, and resource allocation. On the one hand, real-time acquisition, cleaning, and correlation of multi-source data present technical challenges, making it difficult to guarantee data accuracy and timeliness. On the other hand, traditional decision-making models, based on static rules or simple algorithms, have poor adaptability, resulting in low resource allocation efficiency and unsatisfactory treatment outcomes. Furthermore, emergency response lacks real-time monitoring and dynamic adjustment mechanisms for resource execution status, making it difficult to cope with sudden demands for resource reallocation.

[0023] In summary, developing a big data-based medical emergency response method and system is crucial. This method and system should possess functions such as real-time integration and analysis of multi-source data, intelligent injury classification and prediction, and optimized resource scheduling and dynamic adjustment, to significantly improve the speed and accuracy of medical emergency response and better cope with complex and ever-changing emergency scenarios.

[0024] To address the aforementioned issues, this embodiment provides a big data-based medical emergency response method. Through multimodal signal reception, data integration, intelligent injury assessment, optimized resource scheduling, and dynamic adjustment, it achieves rapid, accurate, and efficient medical emergency response, thereby improving the overall effectiveness of emergency handling.

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] Please see Figure 1The diagram shows a flowchart of a big data-based medical emergency response method in a specific embodiment. The method includes the following steps: S1. Receive and respond to medical emergency event signals in real time through a multimodal input channel, classify and grade medical emergency events, and generate structured event description data for medical emergency events; It should be noted that by receiving and responding to medical emergency event signals in real time through the multimodal input channel, it is possible to quickly perceive various forms of emergency event reports, classify and classify events, and generate structured event description data, providing timely and accurate event information for subsequent emergency handling and facilitating the rapid initiation of emergency response procedures. S2. Dynamically collect historical case data, medical resource status data, environmental data, and vital sign assessment data of the injured, and clean and correlate them in real time through stream processing to construct a three-dimensional data map of medical emergency events, medical resources, and environment; It should be noted that dynamic collection of multi-source data, followed by real-time cleaning and correlation through stream processing, constructs a three-dimensional data map that integrates information on the wounded, medical resources, and the environment, providing a data foundation for subsequent analysis and decision-making. S3. Use a gradient boosting tree model to classify the severity of the injuries, combine the medical image analysis results generated by the CNN model to predict the trend of the disease development, and output an injury classification report. It should be noted that by using a gradient boosting tree model to classify the severity of the injured, combining the medical image analysis results generated by the CNN model to predict the trend of the disease, and making corrections and supplements based on the three-dimensional data map, it is possible to more accurately assess the severity and trend of the injured's condition, and provide support for the rational allocation of medical resources and the formulation of treatment plans. S4. Based on the improved VRPTW model and genetic algorithm model, with emergency priority as the constraint and hospital load balancing as the goal, a scheduling scheme is generated from the initial assignment of ambulances to route planning and then to the pre-allocation of medical resources. The scheme is dynamically adjusted as the ambulance location data is updated, and optimization is performed using a three-dimensional data map. It should be noted that, based on the improved VRPTW model and genetic algorithm model, with emergency priority as the constraint and hospital load balancing as the goal, a scheduling scheme is generated and optimized using a three-dimensional data map, which realizes efficient scheduling and optimized allocation of medical resources, ensuring that critically injured patients can receive timely treatment, while avoiding excessive concentration or idleness of hospital resources. S5. Track the execution status of resources in real time, and when the response time exceeds the set range of the expected response time, start the reinforcement learning model to dynamically adjust the scheduling scheme based on the three-dimensional data map and generate an effect evaluation report. It should be noted that by tracking the execution status of resources in real time, when the response time exceeds expectations, the reinforcement learning model is activated to dynamically adjust the scheduling scheme based on the three-dimensional data map and generate an effect evaluation report. This can promptly respond to various changes and emergencies in the emergency response process, ensure the rationality and effectiveness of the scheduling scheme, and improve the dynamic adaptability of emergency response.

[0027] This embodiment achieves rapid perception and classification of medical emergency events through multimodal signal reception, dynamically collects multi-source data and processes it in real time to construct a three-dimensional data map, accurately assesses injuries by combining gradient boosting tree model and CNN model, optimizes resource scheduling scheme by using improved VRPTW model and genetic algorithm, and achieves dynamic adjustment and effect evaluation through reinforcement learning.

[0028] Furthermore, as a refinement and extension of the specific implementation methods of the above embodiments, in order to fully illustrate the specific implementation process of this embodiment, another big data-based medical emergency response method is provided, which includes the following steps: S1. Receive and respond to medical emergency event signals in real time through a multimodal input channel, classify and grade the events, and generate structured event description data; the specific steps of step S1 are as follows: S11. Parse emergency call recordings through a voice recognition interface and extract key fields including the location of the incident, the number of casualties, and the description of the injuries; S12. Connect to roadside intelligent equipment to obtain video streams of the accident scene, and use the YOLOv5 model to identify the distribution density of the injured; S13. Match event types according to the preset rule base, and classify medical emergency events according to the number of casualties, whether there are critically injured people, and whether they involve the agreed type of injury; For example, a Level I response is triggered when there are ≥3 casualties, critically injured patients, or special injuries (chemical / radiation). S14. Generate structured event codes containing geographic location coordinates, timestamps, and medical emergency event classifications; S2. Dynamically collect historical case data, medical resource status data, environmental data, and vital sign assessment data of the injured, and perform real-time cleaning and correlation through stream processing to construct a three-dimensional data map of events, resources, and environment; the specific steps of step S2 are as follows: S21. Obtain medical resource status data from the hospital HIS system and emergency center dispatch system via HTTP / HTTPS interface; the medical resource status data includes real-time hospital bed occupancy, initial location of ambulance, real-time location of ambulance, on-duty status of medical staff, and professional skill tags; The initial location of the ambulance refers to the location obtained from the emergency center dispatch system before dispatch, while the real-time location of the ambulance refers to the location transmitted back via the ambulance's GPS after dispatch. S22. Obtain meteorological warning data through the meteorological bureau's API interface, obtain real-time traffic flow through the map application interface, and use the meteorological warning data and real-time traffic flow as environmental data; S23. Obtain historical case data from a cross-institutional medical data center, wherein the historical case data includes resource scheduling schemes and treatment effect statistics for similar events; S24. Determine vital sign assessment data for the injured based on structured event description data; It should be noted that the vital signs assessment data of the injured are extracted from the description of injury and death in the emergency call recording, as well as the injured person's posture and infrared temperature measurement identification from the video stream collected from the roadside equipment. If the injured person is wearing a smart bracelet or emergency bracelet, heart rate and blood oxygen data can be collected from the corresponding bracelet. The vital signs assessment data of the injured are supplemented after the ambulance arrives at the scene. Parameters such as heart rate and blood oxygen are automatically collected by the on-board monitor or the preliminary triage results are manually entered by medical staff through a PAD. S25. Use the Apache Flink framework to perform real-time cleaning of multi-source data, filter invalid characters through regular expressions, remove abnormal device data based on thresholds, and establish multi-dimensional association relationships between medical emergency events, medical resources, and the environment through association rules to generate a three-dimensional data map containing spatial coordinates, time series, and resource attributes. Specifically, spatial data includes the location of the injured (e.g., coordinates of the accident scene), the location of the hospital (e.g., coordinates of hospital A), and the location of the ambulance (e.g., real-time GPS coordinates). Time data includes the time of the incident, the estimated arrival time of the ambulance, and the time required for hospital resource preparation. Resource data includes the number of ICU beds in hospital A, the equipment configuration of ambulance B (such as ventilators and defibrillators), and the professional skills of medical staff; Association rules include spatial association rules, temporal association rules, and resource association rules; Specific spatial association rules include: the distance between the injured person's location and the nearest hospital, and the real-time distance between the injured person's location and the ambulance; Specific examples of time-related rules include: time-series changes in the vital signs data of the injured, and matching the arrival time of ambulances with the preparation time of hospital resources; Specific resource association rules include: matching the patient's condition with the hospital department (e.g., matching a fracture patient with the trauma surgery department), and matching ambulance resources with the patient's needs (e.g., using a ventilator for a patient with breathing difficulties). S3. A gradient boosting tree model is used to classify the severity of the injured patient's condition. This is combined with medical image analysis results generated using a CNN model to predict the progression of the condition. Further corrections and supplements are made based on 3D data atlases, and a severity classification report is output. The specific steps of step S3 are as follows: S31. Input the patient's vital signs assessment data, patient's age, and basic medical history into the gradient boosting model to determine the severity level of the patient's condition; For example, output a severity score of 0-5, where a score of 4-5 indicates a critically injured patient who should be prioritized for ICU resource allocation; S32. Use a CNN model to identify and classify lesions in the CT / MRI images of the injured, output the lesion probability value, and combine environmental factors in the three-dimensional data map to correct the disease prediction results and generate a disease severity classification report that includes a warning of deterioration risk. Environmental factors such as rescue delays; S33. Employ a spatiotemporal graph neural network to fuse vital sign time-series data and geospatial data to dynamically adjust the priority of medical resource matching corresponding to injury severity levels; For example, consider the distance to nearby hospitals and the availability of their departments; S4. Based on the improved VRPTW model and genetic algorithm model, with emergency priority as the constraint and hospital load balancing as the objective, a scheduling scheme from ambulance dispatch to route planning and pre-allocation of medical resources is generated, and optimized using a three-dimensional data map; the specific steps of step S4 are as follows: S41. Based on injury severity reports, construct an emergency medical resource scheduling model with time windows, wherein: The golden treatment window for critically injured patients and the matching degree of specialists must meet the preset threshold as hard constraints. The optimization objective is to minimize the average response time and maximize the utilization rate of medical resources. S42. An improved Dijkstra algorithm is used to calculate the optimal path for a single wounded soldier. Real-time traffic data is introduced to dynamically update the edge weights, and the initial matching of wounded soldier and hospital is completed using the Hungarian algorithm. S43. Use the ambulance allocation scheme as the first gene segment and the path sequence as the second gene segment for chromosome encoding, and set the crossover probability and mutation probability of the genetic algorithm; S44. Monitor changes in medical resources and traffic data in the 3D data map in real time, and when the trigger threshold is met, retain the feasible region of the current optimal solution and start the local search algorithm for rapid adjustment; For example, hospital bed occupancy rates fluctuate by more than 15%, and road traffic speed decreases by more than 20%; S45. Generate an executable set of dispatch instructions, including an ambulance dispatch list, dynamic navigation routes, and hospital resource readiness notifications; For example, the ambulance dispatch list includes vehicle ID-patient ID-target hospital ID, the dynamic navigation route also includes alternative routes and expected communication time, and the hospital resource preparation notification includes a list of departments and equipment that need to be reserved; S5. Track resource execution status in real time, and when the response time exceeds the set range of the expected response time, activate the reinforcement learning model to dynamically adjust the scheduling scheme based on the 3D data map and generate an effect evaluation report; the specific steps of step S5 are as follows: S51. Real-time collection of multi-dimensional monitoring indicators, including resource status data, casualty status data, and environmental update data; S52. Calculate the core deviation of each monitoring indicator, and initiate root cause analysis when the core deviation exceeds the set threshold; For example, if road congestion leads to: triggering the dynamic optimization in step S44; If the hospital refuses to accept the patient, adjust the matching rules in steps S42 and S43. S53. Constructing dynamic policy optimization for Markov decision processes: Construct the state space S: S = {Injury severity rating, hospital load, road condition rating, remaining time window percentage}; Constructing action space A: A = {Reassign hospital, increase resources, change route, activate emergency protocol}; For example, reassigning a hospital can be rematching a backup hospital (matching degree α≥0.6), the resource gain can be calling up social emergency vehicles, changing the route can be switching to an alternative route or enabling emergency lane permissions, and activating the emergency protocol can be notifying the target hospital to activate the "overcapacity reception plan" (such as temporary bed addition). Construct the reward function:

[0029] in, The actual response time, for example, in minutes; the score is the treatment outcome score. and For the pre-set weights, and ;For example , ; Among them, the treatment outcome score is dynamically calculated based on the stability of the injured person's vital signs when they arrive at the hospital; S54. Use historical scheduling logs and the PPO algorithm to train the agent online, and update the policy network parameters once every set scheduling cycle; S55. Evaluate the effectiveness of emergency response processes based on timeliness indicators, medical quality indicators, and resource efficiency indicators.

[0030] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0031] like Figure 2 As shown, the following are embodiments of the big data-based medical emergency response system provided in this disclosure. This system and the big data-based medical emergency response methods in the above embodiments belong to the same inventive concept. For details not described in detail in the embodiments of the big data-based medical emergency response system, please refer to the embodiments of the big data-based medical emergency response methods described above.

[0032] The system includes: The emergency response module is used to receive and respond to medical emergency signals in real time through a multimodal input channel, classify and classify medical emergency events, and generate structured event description data for medical emergency events. The multi-source data acquisition module is used to dynamically collect historical case data, medical resource status data, environmental data, and vital sign assessment data of the injured. It also performs real-time cleaning and correlation through stream processing to construct a three-dimensional data map of medical emergency events, medical resources, and the environment. The injury analysis module is used to classify the severity of the injuries using a gradient boosting tree model, combine the medical image analysis results generated by the CNN model to predict the development trend of the injury, and output an injury classification report. The resource allocation and scheduling module is used to generate a scheduling scheme based on the improved VRPTW model and genetic algorithm model, with emergency priority as the constraint and hospital load balancing as the goal, from the initial assignment of ambulances to route planning and then to the pre-allocation of medical resources. It is dynamically adjusted as the ambulance location data is updated, and optimized using three-dimensional data maps. The scheduling result evaluation module is used to track the resource execution status in real time. When the response time exceeds the set range of the expected response time, the reinforcement learning model is activated to dynamically adjust the scheduling scheme based on the three-dimensional data map and generate an effect evaluation report.

[0033] This embodiment improves the efficiency and effectiveness of emergency response through the interactive collaboration of the emergency event response module, multi-source data acquisition module, injury analysis module, resource allocation and scheduling module, and scheduling result evaluation module, making it easier to promote and apply in actual medical emergency scenarios.

[0034] The big data-based medical emergency response method provided in this application can be applied to electronic devices. Those skilled in the art will understand that the electronic device structure involved in the embodiments of this invention does not constitute a limitation on the electronic device. An electronic device may include more or fewer components than illustrated, or combine certain components, or have different component arrangements. In the embodiments of this invention, the electronic device includes, but is not limited to, laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the embodiments of this application described and / or claimed herein.

[0035] Electronic devices may include processors, external memory interfaces, internal memory, universal serial bus (USB) interfaces, charging management modules, power management modules, batteries, wireless communication modules, audio modules, speakers, microphones, sensor modules, buttons, cameras, displays, and SIM card interfaces, etc.

[0036] It is understood that the structures illustrated in the embodiments of this application do not constitute a specific limitation on the electronic device. In other embodiments of this application, the electronic device may include more or fewer components than illustrated, or combine some components, or split some components, or have different component arrangements. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

[0037] A processor may include one or more processing units, such as a central processing unit (CPU), an application processor (AP), a modem processor, a graphics processing unit (GPU), an image signal processor (ISP), a controller, memory, a video codec, a digital signal processor (DSP), a baseband processor, and / or a neural network processing unit (NPU). Different processing units may be independent devices or integrated into one or more processors.

[0038] The processor can serve as the nerve center and command center of an electronic device. The controller can generate operation control signals based on the instruction opcode and timing signals to control the fetching and execution of instructions.

[0039] The processor may also include memory for storing instructions and data. In some embodiments, the memory in the processor is a cache memory. This memory can store instructions or data that the processor has just used or that are used repeatedly. If the processor needs to use the instruction or data again, it can retrieve it directly from this memory. This avoids repeated accesses, reduces processor latency, and thus improves system efficiency.

[0040] The aforementioned electronic equipment enables the medical emergency response method based on big data proposed in this application to receive and respond to medical emergency event signals in real time through a multimodal input channel, classify and grade medical emergency events, and generate structured event description data for medical emergency events; dynamically collect historical case data, medical resource status data, environmental data, and vital sign assessment data of the injured, and perform real-time cleaning and correlation through stream processing to construct a three-dimensional data map of medical emergency events, medical resources, and the environment; use a gradient boosting tree model to classify the severity of the injured's condition, combine the medical image analysis results generated by the CNN model to predict the trend of the condition's development, and output a severity classification report; and is based on an improved VRPTW. The model and genetic algorithm model, with emergency priority as a constraint and hospital load balancing as the goal, generate a scheduling scheme from initial ambulance assignment to route planning and pre-allocation of medical resources. The scheme is dynamically adjusted as ambulance location data is updated, and optimized using 3D data maps. The model tracks resource execution status in real time, and when the response time exceeds the set range of the expected response time, a reinforcement learning model is activated to dynamically adjust the scheduling scheme based on the 3D data map and generate an effect evaluation report. This technical solution achieves the beneficial effects of rapid, accurate and efficient medical emergency response through multimodal signal reception, data integration, intelligent injury assessment, optimized resource scheduling and dynamic adjustment, thereby improving the overall effectiveness of emergency treatment.

[0041] The storage medium provided in this application stores a program product capable of implementing a big data-based medical emergency response method.

[0042] The big data-based medical emergency response method includes: receiving and responding to medical emergency event signals in real time through multimodal input channels, classifying and grading medical emergency events, and generating structured event description data for medical emergency events; dynamically collecting historical case data, medical resource status data, environmental data, and patient vital sign assessment data, and cleaning and associating them in real time through stream processing to construct a three-dimensional data map of medical emergency events, medical resources, and the environment; using a gradient boosting tree model to grade the severity of patients' conditions, combining the medical image analysis results generated by the CNN model to predict the trend of disease development, and outputting a severity grading report; based on an improved VRPTW model and a genetic algorithm model, using emergency priority as a constraint and hospital load balancing as an objective, generating a scheduling scheme from initial ambulance assignment to route planning and pre-allocation of medical resources, dynamically adjusting it as ambulance location data is updated, and optimizing it using the three-dimensional data map; tracking resource execution status in real time, and when the response time exceeds the set range of the expected response time, activating a reinforcement learning model to dynamically adjust the scheduling scheme based on the three-dimensional data map, and generating an effect evaluation report.

[0043] In some possible implementations, the big data-based medical emergency response method of this disclosure can be implemented as a program product comprising program code that, when the program product is run on a terminal device, causes the terminal device to perform the steps described in the "Exemplary Methods" section of this specification according to various exemplary embodiments of this disclosure.

[0044] The storage medium disclosed herein may be any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (a non-exhaustive list) of readable storage media include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof.

[0045] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A medical emergency response method based on big data, characterized in that, Includes the following steps: S1. Receive and respond to medical emergency event signals in real time through a multimodal input channel, classify and grade medical emergency events, and generate structured event description data for medical emergency events; S2. Dynamically collect historical case data, medical resource status data, environmental data, and vital sign assessment data of the injured, and clean and correlate them in real time through stream processing to construct a three-dimensional data map of medical emergency events, medical resources, and environment; S3. Use a gradient boosting tree model to classify the severity of the injuries, combine the medical image analysis results generated by the CNN model to predict the trend of the disease development, and output an injury classification report. S4. Based on the improved VRPTW model and genetic algorithm model, with emergency priority as the constraint and hospital load balancing as the goal, a scheduling scheme is generated from the initial assignment of ambulances to route planning and then to the pre-allocation of medical resources. The scheme is dynamically adjusted as the ambulance location data is updated, and optimization is performed using a three-dimensional data map. S5. Track the execution status of resources in real time, and when the response time exceeds the set range of the expected response time, start the reinforcement learning model to dynamically adjust the scheduling scheme based on the three-dimensional data map and generate an effect evaluation report.

2. The medical emergency response method based on big data according to claim 1, characterized in that, The specific steps of step S1 are as follows: S11. Parse emergency call recordings through a voice recognition interface and extract key fields including the location of the incident, the number of casualties, and the description of the injuries; S12. Connect to roadside intelligent equipment to obtain video streams of the accident scene, and use the YOLOv5 model to identify the distribution density of the injured; S13. Match event types according to the preset rule base, and classify medical emergency events according to the number of casualties, whether there are critically injured people, and whether they involve the agreed type of injury; S14. Generate structured event codes that include geographic location coordinates, timestamps, and medical emergency event classifications.

3. The medical emergency response method based on big data according to claim 2, characterized in that, The specific steps of step S2 are as follows: S21. Obtain medical resource status data from the hospital HIS system and emergency center dispatch system via HTTP / HTTPS interfaces; S22. Obtain meteorological warning data through the meteorological bureau's API interface, obtain real-time traffic flow through the map application interface, and use the meteorological warning data and real-time traffic flow as environmental data; S23. Obtain historical case data from a cross-institutional medical data center, wherein the historical case data includes resource allocation plans and treatment effect statistics for similar medical emergency events; S24. Determine vital sign assessment data for the injured based on structured event description data; S25. Use the Apache Flink framework to perform real-time cleaning of multi-source data. Filter invalid characters using regular expressions, remove abnormal device data based on thresholds, and establish multi-dimensional relationships between medical emergency events, medical resources, and the environment through association rules to generate a three-dimensional data map containing spatial coordinates, time series, and resource attributes.

4. The medical emergency response method based on big data according to claim 3, characterized in that, The medical resource status data includes real-time hospital bed occupancy, initial ambulance location, real-time ambulance location, on-duty status of medical staff, and professional skill tags.

5. The medical emergency response method based on big data according to claim 3, characterized in that, The specific steps of step S3 are as follows: S31. Input the patient's vital signs assessment data, patient's age, and basic medical history into the gradient boosting model to determine the severity level of the patient's condition; S32. Use a CNN model to identify and classify lesions in the CT / MRI images of the injured, output the lesion probability value, and generate a severity rating report that includes a warning of the risk of deterioration; S33. Use a spatiotemporal graph neural network to fuse time-series data of vital signs and geospatial data to dynamically adjust the priority of medical resource matching corresponding to injury severity levels.

6. The medical emergency response method based on big data according to claim 5, characterized in that, The specific steps of step S4 are as follows: S41. Based on injury severity reports, construct an emergency medical resource scheduling model with time windows, wherein: The golden treatment window for critically injured patients and the matching degree of specialists must meet the preset threshold as hard constraints. The optimization objective is to minimize the average response time and maximize the utilization rate of medical resources. S42. An improved Dijkstra algorithm is used to calculate the optimal path for a single wounded soldier. Real-time traffic data is introduced to dynamically update the edge weights, and the initial matching of wounded soldier and hospital is completed using the Hungarian algorithm. S43. Use the ambulance allocation scheme as the first gene segment and the path sequence as the second gene segment for chromosome encoding, and set the crossover probability and mutation probability of the genetic algorithm; S44. Monitor changes in medical resources and traffic data in the 3D data map in real time, and when the trigger threshold is met, retain the feasible region of the current optimal solution and start the local search algorithm for rapid adjustment; S45. Generate a set of executable dispatch instructions, including an ambulance dispatch list, dynamic navigation routes, and hospital resource readiness notifications.

7. The medical emergency response method based on big data according to claim 6, characterized in that, The specific steps of step S5 are as follows: S51. Real-time collection of multi-dimensional monitoring indicators, including resource status data, casualty status data, and environmental update data; S52. Calculate the core deviation of each monitoring indicator, and initiate root cause analysis when the core deviation exceeds the set threshold; S53. Constructing dynamic policy optimization for Markov decision processes: Construct the state space S: S = {Injury severity rating, hospital load, road condition rating, remaining time window percentage}; Constructing action space A: A = {Reassign hospital, increase resources, change route, activate emergency protocol}; Construct the reward function: in, The actual response time is represented by the actual response time, and the score represents the treatment outcome score. and For the pre-set weights, and ; Among them, the treatment outcome score is dynamically calculated based on the stability of the injured person's vital signs when they arrive at the hospital; S54. Use historical scheduling logs and the PPO algorithm to train the agent online, and update the policy network parameters once every set scheduling cycle; S55. Evaluate the effectiveness of emergency response processes based on timeliness indicators, medical quality indicators, and resource efficiency indicators.

8. A medical emergency response system based on big data, characterized in that, include: The emergency response module is used to receive and respond to medical emergency signals in real time through a multimodal input channel, classify and classify medical emergency events, and generate structured event description data for medical emergency events. The multi-source data acquisition module is used to dynamically collect historical case data, medical resource status data, environmental data, and vital sign assessment data of the injured. It also performs real-time cleaning and correlation through stream processing to construct a three-dimensional data map of medical emergency events, medical resources, and the environment. The injury analysis module is used to classify the severity of the injuries using a gradient boosting tree model, combine the medical image analysis results generated by the CNN model to predict the development trend of the injury, and output an injury classification report. The resource allocation and scheduling module is used to generate a scheduling scheme based on the improved VRPTW model and genetic algorithm model, with emergency priority as the constraint and hospital load balancing as the goal, from the initial assignment of ambulances to route planning and then to the pre-allocation of medical resources. It is dynamically adjusted as the ambulance location data is updated and optimized using three-dimensional data maps. The scheduling result evaluation module is used to track the resource execution status in real time. When the response time exceeds the set range of the expected response time, the reinforcement learning model is activated to dynamically adjust the scheduling scheme based on the three-dimensional data map and generate an effect evaluation report.

9. An electronic device, characterized in that, The device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the program to implement the steps of the big data-based medical emergency response method as described in any one of claims 1 to 7.

10. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the big data-based medical emergency response method as described in any one of claims 1 to 7.

Images