First-aid method and system based on environmental adaptability

Abstract

The embodiment of the invention provides a first-aid method and system based on environmental adaptability. The method comprises the following steps: firstly, collecting environment information such as temperature and humidity of an accident occurrence place, evaluating extra risks of the environment to the injury, and generating an environment factor evaluation result; and in combination with vital sign data of the wearable medical device and position information of the latest medical institution provided by the geographic positioning service, a personalized first-aid scheme is generated by using a deep learning algorithm. And then, continuously monitoring environment change by adopting a dynamic programming algorithm, predicting results of different decision paths, and creating an emergency scene simulation platform through a virtual reality technology to carry out rehearsal testing and optimize an emergency strategy. And finally, establishing real-time communication connection between on-site first-aid personnel and professional medical personnel by utilizing a 4G / 5G network, integrating opinions of multiple experts, improving a first-aid strategy, and generating a comprehensive scheme including remote guidance support and on-site first-aid action plans. According to the technical scheme provided by the embodiment of the invention, the first-aid response speed and accuracy are improved.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of intelligent first aid technology, and in particular, to a first aid method and system based on environmental adaptability. Background Art

[0002] In emergency medical rescue, environmental conditions (such as temperature and humidity) have an important impact on the additional risks to the injured and the effectiveness of first aid. For example, in extreme weather or complex geographical conditions, traditional first aid methods may not fully consider these factors, resulting in inefficient or poor treatment results. Therefore, a first aid method is needed that can evaluate environmental factors in real time and combine the vital signs data of the injured. In addition, the location of the nearest medical institution needs to be quickly determined during the first aid process, and a personalized first aid plan needs to be provided to ensure that the most effective treatment measures can be implemented as soon as possible.

[0003] Current first aid methods mainly rely on the experience of first aid personnel on the scene and limited equipment support. The common practice is to collect patient information through simple vital signs monitoring equipment and process it according to preset standard procedures. Although some systems have introduced geolocation services to guide the nearest medical institutions, they usually lack a comprehensive assessment of environmental factors and fail to make full use of advanced data analysis technology to generate personalized first aid plans. In addition, existing communication methods are mostly one-way or have large delays, making it difficult to achieve real-time multi-expert consultation and support.

[0004] However, the existing first aid methods have the following major defects: Traditional first aid methods fail to fully consider the impact of environmental factors such as temperature and humidity at the accident site on the injured, which may lead to improper first aid measures and increase the risk of injury. Most of the existing first aid plans are based on fixed standard procedures, fail to combine individual vital signs and specific environmental conditions, and lack personalization and pertinence. Existing communication methods cannot achieve real-time interaction between on-site first aid personnel and professional medical personnel, which limits the support effect of remote guidance and affects the timeliness and accuracy of decision-making.

[0005] In summary, existing first aid methods have obvious deficiencies in dealing with complex environments and personalized needs, and a new first aid method that can fully consider environmental adaptability is urgently needed to improve the speed and success rate of first aid response. This invention aims to overcome the above shortcomings and provide a more scientific and efficient first aid solution by integrating advanced technologies such as environmental assessment, deep learning analysis, virtual reality simulation and 4G / 5G real-time communication. Summary of the invention

[0006] The embodiments of the present application provide a first aid method and system based on environmental adaptability, so as to solve the problems of slow first aid response and low accuracy in the prior art.

[0007] In a first aspect, an embodiment of the present application provides a first aid method based on environmental adaptability, including:

[0008] Collect environmental information on temperature and humidity at the accident site, assess the additional risks posed by the environment to the injured, and obtain environmental factor assessment results;

[0009] Based on the environmental factor assessment results and the patient's vital sign data obtained from the wearable medical device, combined with the location information of the nearest medical institution provided by the geo-positioning service, a deep learning algorithm is used for analysis and processing to generate a personalized first aid plan;

[0010] Based on the dynamic programming algorithm, the changes in environmental conditions are continuously monitored and analyzed in real time, the results of the personalized first aid plan under different decision paths are predicted, and the first aid scenario simulation platform created by virtual reality technology is rehearsed for the personalized first aid plan. The first aid strategy is optimized and adjusted according to the rehearsal results to obtain the optimized first aid strategy;

[0011] Use 4G / 5G networks to establish real-time communication connections between on-site emergency personnel and professional medical personnel, integrate opinions from multiple experts, improve the optimized emergency strategy, and generate a comprehensive emergency action plan that includes remote guidance support and on-site emergency action plans.

[0012] Optionally, the environmental factor assessment result and the patient's vital sign data obtained from the wearable medical device are combined with the nearest medical institution location information provided by the geo-location service, and analyzed and processed using a deep learning algorithm to generate a personalized first aid plan, including:

[0013] Integrate the environmental factor assessment results with the patient's vital sign data obtained from the wearable medical device to obtain comprehensive first aid input data;

[0014] Based on the nearest medical institution location information provided by the geo-location service, the comprehensive emergency input data is associated and processed, and the geographical location-optimized emergency input data is generated in combination with the influence of the geographical location on the emergency measures;

[0015] Using a convolutional neural network to perform feature extraction and pattern recognition on the geographically optimized first aid input data, and analyzing the optimal combination of first aid measures under different environmental conditions and physiological states to generate a preliminary first aid plan;

[0016] The preliminary first aid plan is matched and adjusted based on reasoning technology of historical first aid cases to generate a personalized first aid plan.

[0017] Optionally, the use of a convolutional neural network to perform feature extraction and pattern recognition on the geographically optimized first aid input data, and to analyze the optimal combination of first aid measures under different environmental conditions and physiological states, to generate a preliminary first aid plan, including:

[0018] Using a convolutional neural network to perform feature extraction processing on the geographically optimized emergency input data to obtain a multi-layer feature representation;

[0019] Using a pooling layer to simplify the multi-layer feature representation to obtain a simplified feature representation;

[0020] The simplified feature representation is transformed using a fully connected layer in combination with an activation function to generate a comprehensive feature vector;

[0021] According to the comprehensive feature vector, a regression algorithm is applied to analyze the optimal combination of first aid measures under different conditions to obtain the probability distribution of first aid measures, and based on the probability distribution, a preliminary first aid plan is generated.

[0022] Optionally, the dynamic programming algorithm is used to continuously monitor and instantly analyze changes in environmental conditions, predict the results of the personalized first aid plan under different decision paths, conduct a rehearsal test on the first aid scenario simulation platform created by virtual reality technology for the personalized first aid plan, and optimize and adjust the first aid strategy according to the rehearsal results to obtain an optimized first aid strategy, including:

[0023] Using a dynamic programming algorithm, the real-time environmental condition changes at the accident site are continuously monitored and analyzed in real time, and the environmental challenges faced by the personalized first aid plan at different time points and under different conditions are predicted to obtain environmental change prediction results;

[0024] According to the environmental change prediction results, combined with the combination of measures in the personalized first aid plan, the Markov decision process model is used for analysis and processing to evaluate the success probability and potential risks of each path, and generate multiple decision paths and expected results;

[0025] A first aid scenario simulation platform created based on virtual reality technology is used to perform rehearsal tests on the multiple decision paths and expected results, verify the actual effects of different decision paths, and obtain rehearsal test feedback information;

[0026] According to the feedback information from the rehearsal test, the personalized first aid plan is optimized and adjusted in a targeted manner to generate an optimized first aid strategy, wherein the adjustment content includes the adjustment of the sequence of first aid steps and the re-planning of resource allocation.

[0027] Optionally, the dynamic programming algorithm is used to continuously monitor and immediately analyze the changes in real-time environmental conditions at the accident site, predict the environmental challenges faced by the personalized first aid plan at different time points and under different conditions, and obtain environmental change prediction results, including:

[0028] Using a sensor network deployed on-site or on a remote server to collect real-time environmental parameters at the accident site, wherein the real-time environmental parameters include temperature, humidity and air quality;

[0029] Based on the real-time environmental parameters, a state model of the environmental conditions at the accident site is constructed using the state space representation method in the dynamic programming algorithm, wherein the state model is used to describe the changing rules of the environmental conditions over time and space, and generate an environmental state model;

[0030] Using the value iteration method in the dynamic programming algorithm and the environmental state model, the changing trend of environmental conditions is predicted, and the environmental challenges faced by the personalized first aid program at different time points and under different conditions are predicted, so as to obtain the time series prediction results of environmental changes;

[0031] Based on the time series prediction results of the environmental changes, key environmental factors and change trends that have a significant impact on emergency operations are identified, and environmental change prediction results are generated.

[0032] Optionally, the environmental change prediction result is combined with the combination of measures in the personalized first aid plan, and a Markov decision process model is used for analysis and processing to evaluate the success probability and potential risks of each path, and generate multiple decision paths and expected results, including:

[0033] Integrate the environmental change prediction result and the measure combination in the personalized first aid plan to obtain comprehensive first aid decision input data;

[0034] Based on the Markov decision process model, the state space definition processing is performed on the comprehensive emergency decision input data, all emergency actions are identified as state transitions, and the probability distribution of taking different actions under each state is determined to generate an emergency action state transition diagram;

[0035] By using the emergency action state transition diagram and combining the value iteration algorithm of the Markov decision process model, each emergency path is evaluated for value, and a value evaluation result of each path is generated, wherein the value evaluation content includes the success probability, required time, resource consumption and potential risks of the path;

[0036] According to the value assessment results of each path, the expected result after each path prediction execution is determined, and multiple decision paths and expected results are generated.

[0037] Optionally, the 4G / 5G network is used to establish a real-time communication connection between on-site emergency personnel and professional medical personnel, and the opinions of multiple experts are integrated to improve the optimized emergency strategy, and generate a comprehensive emergency action plan including remote guidance support and on-site emergency action plan, including:

[0038] Using 4G / 5G network technology, establish real-time communication connection between on-site emergency personnel and professional medical personnel to obtain a stable communication link;

[0039] Utilizing the stable communication link, the patient's vital sign data, environmental factor assessment results, and personalized first aid plans are transmitted to medical experts to obtain expert reference information from the medical experts;

[0040] Integrate the opinions of multiple medical experts in different fields, based on the expert reference information, and combine the emergency scenario simulation platform created with virtual reality technology to simulate and adjust the optimized emergency strategy to form a collective wisdom improvement plan;

[0041] Formulate an on-site emergency action plan based on the collective wisdom improvement plan;

[0042] Based on the collective wisdom improvement plan and the on-site first aid action plan, a comprehensive first aid action plan including remote guidance support and on-site first aid action plan is generated.

[0043] In a second aspect, an embodiment of the present application provides a first aid system based on environmental adaptability, including:

[0044] The collection module is used to collect environmental information such as temperature and humidity at the accident site, evaluate the additional risk of the environment to the injured, and obtain the environmental factor evaluation results;

[0045] An analysis module, for generating a personalized first aid plan by using a deep learning algorithm to analyze and process the environmental factor assessment results and the patient's vital sign data obtained from the wearable medical device in combination with the nearest medical institution location information provided by the geolocation service;

[0046] A prediction module is used to continuously monitor and instantly analyze changes in environmental conditions based on a dynamic programming algorithm, predict the results under different decision paths, conduct a rehearsal test on the first aid scenario simulation platform created by virtual reality technology for the personalized first aid plan, and optimize and adjust the first aid strategy according to the rehearsal results to obtain an optimized first aid strategy;

[0047] The integration module is used to establish a real-time communication connection between on-site emergency personnel and professional medical personnel using the 4G / 5G network, integrate the opinions of multiple experts to form collective wisdom, further improve the optimized emergency strategy, and generate a comprehensive emergency action plan that includes remote guidance support and on-site emergency action plan.

[0048] In a third aspect, an embodiment of the present application provides a computing device, comprising a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a first aid method based on environmental adaptability as described in the first aspect above.

[0049] In a fourth aspect, an embodiment of the present application provides a computer storage medium storing a computer program, which, when executed by a computer, implements a first aid method based on environmental adaptability as described in the first aspect.

[0050] In an embodiment of the present application, environmental information such as temperature and humidity at the place where the accident occurred is collected, the additional risks caused by the environment to the injured are evaluated, and the environmental factor evaluation results are obtained; based on the environmental factor evaluation results and the injured vital signs data obtained from the wearable medical device, combined with the location information of the nearest medical institution provided by the geo-positioning service, a deep learning algorithm is used for analysis and processing to generate a personalized first aid plan; based on the dynamic programming algorithm, the changes in environmental conditions are continuously monitored and analyzed in real time, the results of the personalized first aid plan under different decision paths are predicted, and the personalized first aid plan is rehearsed on a first aid scenario simulation platform created by virtual reality technology, and the first aid strategy is optimized and adjusted according to the rehearsal results to obtain an optimized first aid strategy; a real-time communication connection is established between on-site first aid personnel and professional medical personnel using a 4G / 5G network, and the opinions of multiple experts are integrated to improve the optimized first aid strategy to generate a comprehensive first aid action plan including remote guidance support and on-site first aid action plan.

[0051] The technical solution of this application has the following beneficial effects:

[0052] Through real-time environmental monitoring and data analysis, this application can quickly assess the impact of the environment on the injured and quickly generate personalized first aid plans. Use dynamic programming algorithms and virtual reality technology to preview the results of different decision-making paths to ensure the optimization and scientific nature of the first aid strategy. With the help of 4G / 5G networks, real-time communication between on-site first aid personnel and professional medical personnel can be achieved, and the opinions of multiple experts can be integrated to provide instant remote guidance and support. Comprehensively consider environmental factors, vital signs and geographic location information to formulate a more comprehensive and effective first aid action plan, thereby improving the success rate of treatment. Using virtual reality technology to simulate first aid scenarios provides first aid personnel with a more realistic training environment and improves their emergency response capabilities.

[0053] Furthermore, the embodiment of the present application also integrates the environmental factor assessment results with the patient's vital signs data obtained by the wearable medical device to obtain comprehensive first aid input data; based on the location information of the nearest medical institution provided by the geolocation service, the comprehensive first aid input data is associated and processed, and the first aid input data optimized by the geographical location is generated in combination with the influence of the geographical location on the first aid measures. The convolutional neural network is used to extract features and recognize patterns of the first aid input data optimized by the geographical location, and the optimal combination of first aid measures under different environmental conditions and physiological states is analyzed to generate a preliminary first aid plan, and the preliminary plan is matched and adjusted through the reasoning technology of historical first aid cases, and finally a personalized first aid plan is generated. In addition, a dynamic programming algorithm is used to continuously monitor and instantly analyze the real-time environmental changes at the accident site, predict the results of personalized first aid plans under different decision paths, and combine the Markov decision process model to evaluate the success probability and potential risks of each path, and generate multiple decision paths and expected results. The first aid scenario simulation platform created by virtual reality technology previews the actual effects of these paths, and optimizes and adjusts the sequence of first aid steps and resource allocation according to the feedback information, and finally generates an optimized first aid strategy.

[0054] Through the above methods, the speed and accuracy of emergency response are significantly improved. By comprehensively considering environmental factors, vital signs and geographic location information, the scientificity and effectiveness of personalized emergency plans are ensured. Dynamic monitoring and real-time analysis of environmental changes, combined with Markov decision process models and virtual reality technology, not only improve the scientificity and flexibility of decision-making, but also enhance the ability to deal with complex situations, thereby greatly improving the success rate of treatment. Real-time communication connects professional medical personnel, further optimizes emergency strategies, ensures that on-site emergency personnel can obtain the most timely and professional guidance and support, and overall improves the effectiveness and efficiency of emergency operations.

[0055] These and other aspects of the present application will become more clearly understood in the description of the following embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0057] Figure 1 A flow chart of a first aid method based on environmental adaptability provided by the present application is shown;

[0058] Figure 2A schematic structural diagram of an environmental adaptability first aid system provided by the present application is shown;

[0059] Figure 3 A schematic diagram of the structure of a computing device provided by the present application is shown. DETAILED DESCRIPTION

[0060] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0061] In some of the processes described in the specification and claims of this application and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.

[0062] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0063] Figure 1 A flowchart of a first aid method based on environmental adaptability is provided for the embodiment of the present application, such as Figure 1 As shown, the method includes:

[0064] 101. Collect environmental information on temperature and humidity at the accident site, assess the additional risk of the environment to the injured, and obtain the environmental factor assessment results;

[0065] Environmental information collection and risk assessment include collecting environmental data such as temperature and humidity at the accident site. These data not only reflect the current weather conditions, but may also reveal other potential environmental factors (such as air quality and light intensity) to assess the additional risks posed by the environment to the injured. For example, high temperatures may cause heat stroke, and humid environments may increase the risk of infection. These data are obtained through sensors or meteorological APIs, and risk assessment is performed in combination with historical data and expert rule bases to generate environmental factor assessment results.

[0066] First, the emergency system collects environmental data such as temperature and humidity in real time through a network of sensors installed at or near the accident site. Then, it uses a pre-trained risk assessment model to analyze this data and identify environmental factors that may affect the health of the injured. The model is based on a large number of historical cases and expert experience and can accurately predict the risk level under different environmental conditions. Finally, the system generates detailed environmental factor assessment results, which provide an important basis for subsequent steps.

[0067] In a traffic accident rescue scenario, the first aid personnel immediately activated the portable environmental monitoring equipment upon arriving at the scene, and uploaded data such as temperature and humidity to the central control system in real time. The system quickly evaluated the data and found that the temperature at the scene was as high as 35°C and the humidity was high, indicating that there might be a risk of heat stroke. Subsequently, the system recommended that the first aid personnel take cooling measures and incorporate this information into the comprehensive first aid plan to ensure that the impact of environmental factors is fully considered during the subsequent treatment process.

[0068] 102. Based on the environmental factor assessment results and the patient's vital signs data obtained from the wearable medical device, combined with the location information of the nearest medical institution provided by the geo-location service, a deep learning algorithm is used for analysis and processing to generate a personalized first aid plan;

[0069] The generation of personalized first aid plans involves integrating the results of environmental factor assessments, the patient's vital signs data (such as heart rate, blood pressure, blood oxygen saturation), and the location information of the nearest medical institution. These multi-source data are input into a deep learning model, which generates personalized first aid plans through feature extraction and pattern recognition, and analyzes the optimal combination of first aid measures under different environmental conditions and physiological states. This not only improves the pertinence of first aid measures, but also optimizes resource allocation to ensure that the most effective treatment methods are implemented.

[0070] The system receives data from environmental monitoring devices and wearable medical devices, and calls geolocation services to obtain the location information of the nearest medical institution. After preprocessing, these data are input into a convolutional neural network for feature extraction and pattern recognition. The convolutional neural network model is trained based on a large number of historical first aid cases and can identify the combination of first aid measures that best suits the current situation. Finally, the system generates a preliminary first aid plan based on the model output, and further matches and adjusts it through reasoning technology to ensure the individuality and scientificity of the plan.

[0071] Continuing with the aforementioned traffic accident rescue scenario, emergency personnel collected environmental and vital sign data using portable environmental monitoring equipment and smart bracelets worn by the injured. The system combined this data with the location information of the nearest hospital, used deep learning models to analyze and generate a preliminary emergency plan that included cooling, hemostasis, infusion and other measures. The system also recommended that the patient be sent to the nearest emergency hospital as soon as possible, and provided the optimal route planning to ensure that the injured can receive professional treatment in the shortest time.

[0072] 103. Based on the dynamic programming algorithm, continuously monitor and analyze the changes in environmental conditions in real time, predict the results of the personalized first aid plan under different decision paths, conduct a rehearsal test on the first aid scenario simulation platform created by virtual reality technology for the personalized first aid plan, optimize and adjust the first aid strategy according to the rehearsal results, and obtain an optimized first aid strategy;

[0073] Dynamic programming and virtual reality simulation aims to continuously monitor environmental changes and use the Markov decision process model to predict the results of personalized first aid plans under different decision paths. The system uses virtual reality technology to create a realistic first aid scenario simulation platform, conduct rehearsal tests on different paths, and evaluate their actual effects. Based on the rehearsal feedback information, the system can optimize and adjust the sequence of first aid steps and resource allocation in a targeted manner to ensure that the final first aid strategy is both scientific and practical.

[0074] The system uses a dynamic programming algorithm to monitor environmental changes at the accident site in real time and analyze the impact of these changes on the emergency plan in real time. The Markov decision process model is used to predict environmental challenges at different time points and under different conditions, and generate environmental change prediction results. Then, the system uses virtual reality technology to create an emergency scenario simulation platform, conduct rehearsal tests on multiple decision paths, and verify their actual effects. Based on the rehearsal feedback information, the system specifically optimizes and adjusts the sequence of emergency steps and resource allocation to generate an optimized emergency strategy.

[0075] In the traffic accident rescue scenario, as the rescue work unfolds, environmental conditions change, the temperature drops, and the wind increases. The system updates environmental data in real time through a dynamic programming algorithm, and uses a Markov decision process model to predict the impact of these changes on existing first aid plans. The system simulates multiple decision paths on a virtual reality platform, including different transfer methods and schedules. The preview results show that the original transfer route is greatly affected by the weather. The system immediately adjusted the plan, selected a safer alternative route, and replanned resource allocation to ensure that the injured can receive treatment under the best conditions.

[0076] 104. Use 4G / 5G networks to establish real-time communication connections between on-site emergency personnel and professional medical personnel, integrate opinions from multiple experts, improve the optimized emergency strategy, and generate a comprehensive emergency action plan that includes remote guidance support and on-site emergency action plans.

[0077] Real-time communication and multi-party collaboration use 4G / 5G networks to establish real-time communication connections between on-site emergency personnel and remote professional medical personnel, ensuring that both parties can communicate and share information instantly. The system integrates the opinions of multiple experts, further improves the optimized emergency strategy through remote guidance and support, and finally generates a comprehensive emergency action plan that includes remote guidance support and on-site emergency action plan. This approach not only improves the scientificity and timeliness of decision-making, but also enhances the response capabilities of on-site emergency personnel.

[0078] The system enables real-time video calls and data transmission through 4G / 5G networks, allowing on-site emergency personnel to maintain close contact with remote professional medical personnel. Professional medical personnel provide remote guidance and support based on the real-time video images and data transmitted back to help on-site personnel make the best decisions. The system also integrates the opinions of multiple experts and integrates the suggestions of all parties through a collaborative work platform to ensure that the final generated comprehensive emergency action plan is both scientific and feasible. In the traffic accident rescue scenario, the on-site emergency personnel established a real-time communication connection with the emergency doctors in the hospital through the 4G / 5G network. The doctor saw the on-site situation through the video and provided detailed remote guidance based on the vital signs data transmitted back, such as adjusting the infusion speed and selecting appropriate hemostatic materials. At the same time, many experts discussed together through the collaborative work platform and put forward suggestions for further optimization, such as adding psychological counseling to relieve the tension of the injured. In the end, the system generated a detailed comprehensive emergency action plan to ensure that the injured can receive the most effective treatment in the shortest time.

[0079] Through the implementation of steps 101 to 104, the present invention significantly improves the speed and accuracy of emergency response. First, comprehensive consideration of environmental factors, vital signs and geographic location information ensures the scientificity and effectiveness of personalized first aid plans. Secondly, dynamic monitoring and real-time analysis of environmental changes, combined with Markov decision process models and virtual reality technology, not only improves the scientificity and flexibility of decision-making, but also enhances the ability to deal with complex situations. Finally, the 4G / 5G network realizes real-time communication to connect professional medical personnel, further optimizes the first aid strategy, and ensures that on-site first aid personnel can obtain the most timely and professional guidance and support. On the whole, this method greatly improves the success rate and efficiency of treatment, and enhances the intelligence and coordination level of the first aid system.

[0080] Furthermore, in order to improve the efficiency and accuracy of emergency response, the present invention proposes an intelligent and efficient environmental information collection mechanism. The mechanism can flexibly adjust the collection frequency according to actual conditions, which not only ensures the real-time and accuracy of the data, but also effectively saves equipment resources.

[0081] The mechanism includes: Under normal circumstances, the system is set to automatically wake up the module every five minutes to collect environmental information such as temperature and humidity at the accident site. This data is then transmitted to the collection module for local storage and immediately goes back to sleep to save power. This timed triggering method ensures regular updates of environmental information while minimizing unnecessary energy consumption.

[0082] When the wearable medical device detects abnormal fluctuations in the patient's vital signs (such as a sudden increase in heart rate or a sharp drop in blood oxygen saturation), it will immediately wake up the relevant modules to collect data more frequently and report the abnormal status to the central processing unit. At this time, not only the frequency of environmental information collection increases, but also vital sign information will be monitored more intensively to capture any important changes that may affect emergency decision-making in a timely manner.

[0083] Once the device automatically assesses through the built-in algorithm that medical intervention is needed (such as judging that the condition is worsening or there is another emergency), it will immediately initiate a remote guidance request. This process involves using the 4G / 5G network to establish a real-time communication connection between on-site first responders and professional medical personnel, allowing the latter to quickly understand the current situation and provide necessary guidance and support.

[0084] The design of the above mechanism fully considers the energy efficiency management of the device. In non-emergency situations, maintaining a low data collection frequency through timed triggering can effectively extend battery life; in emergency situations, it quickly switches to a high-frequency collection mode to ensure that important information is not missed. In addition, after each collection, the data is immediately transmitted to the collection module for local storage, which facilitates subsequent data analysis and processing, and also reduces the pressure of wireless transmission.

[0085] In the case of abnormal vital signs, the system can respond immediately, which not only improves the speed of emergency response, but also increases the chance of successful treatment. In this way, even in remote areas or under poor communication conditions, key information can be conveyed to the rescue team as soon as possible.

[0086] This solution incorporates an advanced automated evaluation mechanism that can determine when a remote guidance request is needed without human intervention. This intelligent design not only improves the reliability and safety of the system, but also provides a scientific basis for decision-making during the emergency treatment process. A comprehensive analysis is performed using a deep learning algorithm, combining the environmental factor assessment results, vital sign data, and the location information of the nearest medical institution to generate the most optimized personalized emergency treatment plan. The application of a dynamic programming algorithm further enhances the flexibility of the solution, allowing emergency treatment strategies to be adjusted in a timely manner according to changing environmental conditions.

[0087] In summary, this expansion plan significantly improves the practicality and effectiveness of first aid methods based on environmental adaptability by introducing an intelligent environmental information collection mechanism, laying a solid foundation for achieving efficient and accurate first aid services.

[0088] In order to further improve the accuracy and scientificity of the personalized first aid plan, in some embodiments, the step 102 generates a personalized first aid plan by analyzing and processing the environmental factor assessment results and the patient's vital sign data obtained from the wearable medical device in combination with the nearest medical institution location information provided by the geolocation service using a deep learning algorithm, including:

[0089] The environmental factor assessment results are integrated with the patient's vital signs data obtained from the wearable medical device to obtain comprehensive first aid input data; based on the location information of the nearest medical institution provided by the geo-positioning service, the comprehensive first aid input data is associated and processed, and combined with the influence of the geographical location on the first aid measures, the geographical location optimized first aid input data is generated; the convolutional neural network is used to extract features and recognize patterns of the geographical location optimized first aid input data, and the optimal combination of first aid measures under different environmental conditions and physiological states is analyzed to generate a preliminary first aid plan; the preliminary first aid plan is matched and adjusted based on the reasoning technology of historical first aid cases to generate a personalized first aid plan.

[0090] In this embodiment, the comprehensive first aid input data includes the results of environmental factor assessments and the patient's vital signs data (such as heart rate, blood pressure, blood oxygen saturation, etc.), which are used to comprehensively assess the patient's current condition. In addition, the geographically optimized first aid input data refers to the correlation processing of the comprehensive first aid input data based on the location information of the nearest medical institution provided by the geographic positioning service, and the influence of the geographical location on the first aid measures is considered to generate more accurate first aid input data. Convolutional neural networks are used for feature extraction and pattern recognition, and can analyze the optimal combination of first aid measures under different environmental conditions and physiological states. Finally, the first aid plan initially generated is matched and adjusted through the reasoning technology of historical first aid cases to ensure that the personalized first aid plan finally generated is both scientific and practical.

[0091] In an embodiment of the present application, the environmental factor assessment results are first integrated with the patient's vital signs data obtained from the wearable medical device to form comprehensive first aid input data. Then, based on the location information of the nearest medical institution provided by the geo-positioning service, the comprehensive first aid input data is associated and processed, and the geographical location optimized first aid input data is generated in combination with the impact of the geographical location on the first aid measures. Next, a convolutional neural network is used to perform feature extraction and pattern recognition on the geographical location optimized first aid input data, analyze the optimal combination of first aid measures under different environmental conditions and physiological states, and generate a preliminary first aid plan. Finally, based on the reasoning technology of historical first aid cases, the preliminary first aid plan is matched and adjusted to generate a final personalized first aid plan.

[0092] Here is a specific example:

[0093] In a traffic accident rescue scenario in a remote mountainous area, emergency personnel immediately activated portable environmental monitoring equipment upon arrival at the scene, uploading environmental data such as temperature and humidity to the central control system in real time. The system quickly evaluated the data and found that the temperature was low and the humidity was high, indicating the possible risk of hypothermia and infection. At the same time, the smart bracelet worn by the injured continued to transmit vital signs data such as heart rate and blood pressure. The system combined this data with the location information of the nearest hospital to generate geographically optimized emergency input data.

[0094] Subsequently, the system used convolutional neural networks to extract features and recognize patterns from these optimized data, and analyzed the best combination of first aid measures, such as taking immediate warming measures and using antibacterial drugs to prevent infection. The system also referred to similar historical first aid cases, made detailed matches and adjustments, and finally generated a personalized first aid plan that included cooling, hemostasis, infusion and other measures. The system recommends sending the patient to the nearest hospital with emergency capabilities as soon as possible, and provides the best route planning to ensure that the injured can receive professional treatment in the shortest time.

[0095] Through this series of operations, the first aid plan not only takes into account environmental factors and vital signs, but also combines geographic location information, ensuring the scientific nature and pertinence of the plan and improving the success rate and efficiency of treatment.

[0096] In order to further improve the accuracy of feature extraction and pattern recognition and ensure that the generated preliminary first aid plan is more scientific and personalized, in some embodiments, the use of a convolutional neural network in step 102 performs feature extraction and pattern recognition on the geographically optimized first aid input data, and analyzes the optimal combination of first aid measures under different environmental conditions and physiological states to generate a preliminary first aid plan, including:

[0097] A convolutional neural network is used to perform feature extraction processing on the geographically optimized first aid input data to obtain a multi-layer feature representation; a pooling layer is used to simplify the multi-layer feature representation to obtain a simplified feature representation; a fully connected layer is used in combination with an activation function to transform the simplified feature representation to generate a comprehensive feature vector; based on the comprehensive feature vector, a regression algorithm is used to analyze the optimal combination of first aid measures under different conditions to obtain a probability distribution of first aid measures, and a preliminary first aid plan is generated based on the probability distribution.

[0098] In this embodiment, the convolutional neural network is a deep learning model that is particularly good at processing multi-dimensional data, such as images, time series, etc. It extracts features from the input data through multiple convolutional layers to obtain a multi-layer feature representation. The feature map calculated by each convolutional layer can capture the different levels of features of the input data, so as to understand the input information more comprehensively. The pooling layer is used to reduce the computational complexity while maintaining important information and generating a simplified feature representation. The fully connected layer combines the activation function to transform the simplified feature representation and generate a comprehensive feature vector. Finally, the optimal combination of first aid measures under different conditions is analyzed by a regression algorithm to obtain the probability distribution of first aid measures, and a preliminary first aid plan is generated based on this probability distribution.

[0099] In an embodiment of the present application, a convolutional neural network is first used to perform feature extraction processing on geographically optimized first aid input data, and a multi-layer feature representation is calculated through multiple convolutional layers. These feature representations can capture the key features of the injured under different environmental conditions and physiological states. Next, the multi-layer feature representation is simplified using a pooling layer to reduce redundant information while maintaining important features to generate a simplified feature representation. Then, the simplified feature representation is transformed through a fully connected layer combined with an activation function to generate a comprehensive feature vector. Finally, a regression algorithm is applied based on the comprehensive feature vector to analyze the optimal combination of first aid measures under different conditions, obtain the probability distribution of first aid measures, and generate a preliminary first aid plan based on this probability distribution.

[0100] Here is a specific example:

[0101] In a traffic accident rescue scenario in a city center, emergency personnel immediately start portable environmental monitoring equipment after arriving at the scene, and upload environmental data such as temperature and humidity to the central control system in real time. The system quickly evaluates this data and finds that the temperature and humidity at the scene are high, indicating a possible risk of heat stroke. At the same time, the smart bracelet worn by the injured continuously transmits vital signs data such as heart rate and blood pressure. The system combines this data with the location information of the nearest hospital to generate geographically optimized emergency input data.

[0102] Next, the system uses convolutional neural networks to perform feature extraction on these optimized data. The convolutional layers gradually extract features at different levels. For example, the first layer may capture basic changes in vital signs, while deeper layers may identify complex environmental and physiological interactions. Subsequently, the pooling layer simplifies these multi-layer feature representations, reducing computational complexity while retaining key information. For example, the maximum pooling operation can highlight the most significant feature changes, helping the system better understand the current situation.

[0103] Then, the system transforms the simplified feature representation through a fully connected layer combined with an activation function to generate a comprehensive feature vector. This vector contains all the important features that have been processed and simplified, and can fully describe the current environment and the status of the injured. Finally, the system applies a regression algorithm to analyze the optimal combination of first aid measures under different conditions, and obtains the probability distribution of selection of various first aid measures. For example, the probability of selecting cooling measures is higher because the system predicts that the risk of heat stroke is greater in high temperature environments. Based on these probability distributions, the system generates a preliminary first aid plan that includes cooling, hemostasis, infusion and other measures, and recommends that the injured be sent to the nearest hospital with emergency capabilities as soon as possible to ensure that the injured can receive professional treatment in the shortest time.

[0104] Through this series of operations, the first aid plan not only takes into account environmental factors and vital signs, but also combines geographic location information, ensuring the scientific nature and pertinence of the plan and improving the success rate and efficiency of treatment.

[0105] In order to further improve the response speed to environmental changes and the scientific nature of the first aid strategy, in some embodiments, the dynamic programming algorithm in step 103 continuously monitors and instantly analyzes changes in environmental conditions, predicts the results of the personalized first aid plan under different decision paths, and performs a rehearsal test on the first aid scenario simulation platform created by virtual reality technology for the personalized first aid plan, optimizes and adjusts the first aid strategy according to the rehearsal results, and obtains an optimized first aid strategy, including:

[0106] Using a dynamic programming algorithm, the real-time environmental condition changes at the accident site are continuously monitored and analyzed in real time, and the environmental challenges faced by the personalized first aid plan at different time points and under different conditions are predicted to obtain environmental change prediction results; based on the environmental change prediction results, combined with the combination of measures in the personalized first aid plan, the Markov decision process model is used for analysis and processing to evaluate the success probability and potential risks of each path, and generate multiple decision paths and expected results; based on the first aid scenario simulation platform created with virtual reality technology, the multiple decision paths and expected results are rehearsed and tested to verify the actual effects of different decision paths and obtain rehearsal test feedback information; based on the rehearsal test feedback information, the personalized first aid plan is optimized and adjusted in a targeted manner to generate an optimized first aid strategy, wherein the adjustment content includes the sequence adjustment of the first aid steps and the re-planning of resource allocation.

[0107] In this embodiment, a dynamic programming algorithm is used to continuously monitor and instantly analyze changes in real-time environmental conditions at the accident site. The algorithm can effectively capture the time series characteristics of environmental changes, predict the environmental challenges faced by personalized first aid plans at different time points and under different conditions, and generate environmental change prediction results. The Markov decision process model is used to combine the combination of measures in the personalized first aid plan, evaluate the success probability and potential risks of each path, and generate multiple decision paths and expected results. The first aid scenario simulation platform created by virtual reality technology is used to rehearse these paths and expected results, verify the actual effects of different decision paths, and obtain rehearsal test feedback information. Finally, the system optimizes and adjusts the personalized first aid plan in a targeted manner based on this feedback information to ensure that the final generated first aid strategy is more scientific and practical.

[0108] In the embodiment of the present application, the real-time environmental condition changes at the accident site are first continuously monitored and analyzed in real time using a dynamic programming algorithm to generate environmental change prediction results. Then, based on these prediction results, combined with the combination of measures in the personalized first aid plan, the Markov decision process model is used for analysis and processing to evaluate the success probability and potential risks of each path, and to generate multiple decision paths and expected results. Next, a first aid scenario simulation platform created based on virtual reality technology is used to perform a rehearsal test on these multiple decision paths and expected results, to test the actual effects of different decision paths, and to obtain rehearsal test feedback information. Finally, the system optimizes and adjusts the personalized first aid plan in a targeted manner based on these feedback information. The adjustments include adjusting the order of first aid steps and re-planning resource allocation to generate an optimized first aid strategy.

[0109] Here is a specific example:

[0110] In a traffic accident rescue scenario in a city center, the on-site environmental conditions are complex and change rapidly. After arriving at the scene, the emergency personnel immediately started the portable environmental monitoring equipment and uploaded environmental data such as temperature and humidity to the central control system in real time. The system quickly evaluated the data and found that the temperature and humidity at the scene were high, indicating that there might be a risk of heat stroke. At the same time, the smart bracelet worn by the injured continued to transmit vital signs data such as heart rate and blood pressure. The system combined this data with the location information of the nearest hospital to generate geographically optimized emergency input data, and initially generated a personalized emergency plan including cooling, hemostasis, infusion and other measures.

[0111] As the rescue work unfolded, environmental conditions changed, with temperatures dropping and winds picking up. The system used a dynamic programming algorithm to update environmental data in real time, predicting the impact of these changes on existing first aid plans. For example, the system predicted that the casualty might be at risk of hypothermia in a low-temperature environment and would need additional warming measures. The system generated predictions of environmental changes and fed them into a Markov decision process model to assess the success probability and potential risks of different decision paths. The model generated multiple decision paths, such as choosing different modes of transport or adjusting the priority of first aid measures.

[0112] Next, the system used virtual reality technology to create a realistic emergency scenario simulation platform to rehearse these decision paths. For example, one path may involve rapid transfer using a helicopter, while another path uses a ground ambulance. By simulating the actual effects of different paths, the system obtained detailed rehearsal test feedback information. The results showed that the original ground transfer route was greatly affected by the weather. The system immediately adjusted the plan, selected a safer alternative route, and re-planned resource allocation, such as increasing the supply of warming equipment and medicines.

[0113] Finally, the system optimized and adjusted the personalized first aid plan based on the feedback information and generated an optimized first aid strategy. The new strategy not only takes into account the latest environmental changes, but also optimizes the sequence of first aid steps and resource allocation to ensure that the injured can receive treatment under the best conditions. Through this series of operations, the first aid plan not only improves the scientificity and flexibility of decision-making, but also enhances the ability to deal with complex situations, greatly improving the success rate and efficiency of treatment.

[0114] Through the above improvements, the present invention significantly improves the response speed to environmental changes and the scientific nature of the first aid strategy. The continuous monitoring and real-time analysis of the dynamic programming algorithm ensures the timely capture of environmental changes, the evaluation of the Markov decision process model provides a scientific basis for decision-making, and the preview test of virtual reality technology verifies the actual effect of different paths. The optimized first aid strategy finally generated not only improves the accuracy and flexibility of decision-making, but also enhances the intelligence and coordination level of the first aid system, significantly improving the success rate and efficiency of treatment.

[0115] In order to further improve the prediction accuracy of environmental changes and the adaptability of emergency strategies, in some embodiments, the use of a dynamic programming algorithm in step 103 continuously monitors and instantly analyzes changes in real-time environmental conditions at the accident site, predicts environmental challenges faced by the personalized emergency plan at different time points and under different conditions, and obtains environmental change prediction results, including:

[0116] Using a sensor network deployed on-site or on a remote server, real-time environmental parameters of the accident site are collected, wherein the real-time environmental parameters include temperature, humidity and air quality; based on the real-time environmental parameters, a state model of the environmental conditions of the accident site is constructed using the state space representation method in the dynamic programming algorithm, wherein the state model is used to describe the changing rules of environmental conditions over time and space, and to generate an environmental state model; using the value iteration method in the dynamic programming algorithm and the environmental state model, the changing trend of environmental conditions is predicted, and the environmental challenges faced by the personalized first aid plan at different time points and under different conditions are predicted, and a time series prediction result of environmental changes is obtained; based on the time series prediction result of environmental changes, key environmental factors and changing trends that have a significant impact on first aid actions are identified, and environmental change prediction results are generated.

[0117] In this embodiment, the sensor network is deployed on-site or on a remote server to collect real-time environmental parameters such as temperature, humidity and air quality at the accident site. These data are the basis for evaluating the impact of the environment on the injured. State space representation is a mathematical tool used to construct a state model that describes the law of environmental conditions changing over time and space. By abstracting environmental conditions into a series of state variables, the trend of environmental changes can be captured more accurately. The value iteration method is a technique in the dynamic programming algorithm that is used to optimize the decision-making process and predict the changing trend of future environmental conditions by repeatedly iterating and calculating the value function of each state. The final generated environmental change prediction results not only include time series predictions, but also identify key environmental factors and their changing trends that have a significant impact on emergency operations. Key environmental factors mainly include temperature, humidity, air quality, wind speed and direction, precipitation and light intensity.

[0118] In the embodiment of the present application, firstly, a sensor network deployed on site or a remote server is used to collect real-time environmental parameters of the accident site, including temperature, humidity and air quality. Then, based on these real-time environmental parameters, the state space representation method in the dynamic programming algorithm is used to construct a state model of the environmental conditions at the accident site, describe the changing laws of environmental conditions over time and space, and generate an environmental state model. Next, the value iteration method in the dynamic programming algorithm and the environmental state model are used to predict the changing trend of environmental conditions, predict the environmental challenges faced by personalized first aid plans at different time points and under different conditions, and obtain the time series prediction results of environmental changes. Finally, according to the time series prediction results of environmental changes, the key environmental factors and changing trends that have a significant impact on first aid actions are identified, and the environmental change prediction results are generated.

[0119] Here is a specific example:

[0120] In a traffic accident rescue scenario in a city center, the on-site environmental conditions are complex and change rapidly. After arriving at the scene, the emergency personnel immediately activated the portable environmental monitoring equipment and uploaded environmental data such as temperature, humidity, and air quality to the central control system in real time. The system quickly evaluated the data and found that the temperature and humidity at the scene were high, indicating that there might be a risk of heat stroke. At the same time, the air quality was poor and the PM2.5 concentration was high, increasing the risk of respiratory diseases.

[0121] As the rescue work unfolded, environmental conditions changed, the temperature dropped, and the wind increased. The system continuously collected real-time environmental parameters through the sensor network deployed on site, and used the state space representation method to build a state model of the environmental conditions at the accident site. This model can accurately describe the changing laws of environmental conditions over time and space, and generate an environmental state model. Next, the system uses the value iteration method in the dynamic programming algorithm, combined with the environmental state model, to predict the changing trend of environmental conditions. For example, the system predicts that the temperature will continue to drop, the wind speed will increase, and the air quality will improve in the next few hours.

[0122] Based on these predictions, the system identifies key environmental factors and their changing trends that have a significant impact on emergency response. For example, cooling may cause the patient to have hypothermia, requiring additional warming measures; increased wind speed may affect the safety of helicopter transfer, and a ground ambulance is recommended. The system generates detailed predictions of environmental changes and feeds them back to emergency personnel. Based on this information, emergency personnel adjusted the initial emergency plan, increased the supply of warming equipment and medicines, and replanned the transfer route to ensure that the injured can receive professional treatment in the shortest time.

[0123] Through this series of operations, the first aid plan not only takes into account the latest environmental changes, but also optimizes the sequence of first aid steps and resource allocation to ensure that the injured can receive treatment under the best conditions. The final environmental change prediction results improve the scientificity and flexibility of decision-making, enhance the ability to deal with complex situations, and greatly improve the success rate and efficiency of treatment.

[0124] Through the above improvements, the present invention significantly improves the prediction accuracy of environmental changes and the adaptability of emergency response strategies. Real-time data acquisition of sensor networks and state modeling of state space representation ensure timely capture and accurate description of environmental changes, while the prediction of value iteration method provides a scientific basis for decision-making. The environmental change prediction results finally generated not only improve the accuracy and flexibility of decision-making, but also enhance the intelligence and coordination level of the emergency response system, significantly improving the success rate and efficiency of treatment.

[0125] In order to further improve the scientificity and accuracy of emergency path evaluation, in some embodiments, the prediction result of the environmental change in step 103 is combined with the combination of measures in the personalized emergency plan, and the Markov decision process model is used for analysis and processing to evaluate the success probability and potential risks of each path, and generate multiple decision paths and expected results, including:

[0126] The environmental change prediction results and the combination of measures in the personalized first aid plan are integrated to obtain comprehensive first aid decision input data; based on the Markov decision process model, the comprehensive first aid decision input data is processed by state space definition, all first aid actions are identified as state transitions, and the probability distribution of taking different actions in each state is determined to generate a first aid action state transition diagram; using the first aid action state transition diagram, combined with the value iteration algorithm of the Markov decision process model, each first aid path is value evaluated to generate a value evaluation result of each path, wherein the value evaluation content includes the success probability of the path, the required time, the resource consumption and the potential risks; according to the value evaluation results of each path, the expected result after the execution of the prediction of each path is determined, and multiple decision paths and expected results are generated.

[0127] In this embodiment, the comprehensive first aid decision input data includes the prediction results of environmental changes and the combination of measures in the personalized first aid plan. These data are used to comprehensively evaluate the first aid actions and their effects under different environmental conditions. The Markov decision process model is a mathematical framework for modeling decision processes, in which each state can be transferred to other states through a series of possible actions, and each transfer has a corresponding probability distribution. The state space definition process is to abstract all possible first aid actions into a state transition diagram, determine the probability distribution of taking different actions in each state, and thus generate a first aid action state transition diagram. The value iteration algorithm is used to perform value assessment on each first aid path and generate the value assessment results of each path, including the success probability, required time, resource consumption and potential risks of the path.

[0128] In an embodiment of the present application, first, the environmental change prediction results and the combination of measures in the personalized first aid plan are integrated to obtain comprehensive first aid decision input data. Then, based on the Markov decision process model, the comprehensive first aid decision input data is processed by state space definition, all first aid actions are identified as state transitions, and the probability distribution of taking different actions in each state is determined to generate a first aid action state transition diagram. Next, using the first aid action state transition diagram, combined with the value iteration algorithm of the Markov decision process model, each first aid path is value evaluated to generate value evaluation results for each path. Finally, based on these value evaluation results, the expected results after the execution of each path prediction are determined, and multiple decision paths and expected results are generated.

[0129] Here is a specific example:

[0130] In a traffic accident rescue scenario in a city center, the system has obtained detailed prediction results of environmental changes through sensor networks and dynamic programming algorithms, such as temperature drop, wind speed increase, and air quality improvement in the next few hours. At the same time, the system has also initially generated a personalized first aid plan including cooling, hemostasis, infusion and other measures.

[0131] Next, the system integrates these environmental change predictions and the combination of measures in the personalized first aid plan to obtain comprehensive first aid decision input data. For example, the system considers the necessity of cooling measures in low temperature environments and the effectiveness of hemostasis measures in high humidity environments.

[0132] Then, based on the Markov decision process model, the system performs state space definition processing on the comprehensive emergency decision input data. The system identifies all possible emergency actions (such as helicopter transfer, ground ambulance transfer, on-site treatment, etc.) and abstracts them into a state transition diagram. For each state (such as the current condition of the injured, environmental conditions, etc.), the system determines the probability distribution of taking different actions (such as choosing different transfer methods) and generates an emergency action state transition diagram. For example, the system evaluates that the probability of choosing helicopter transfer is higher under certain conditions, while it is more appropriate to choose ground ambulance transfer under other conditions.

[0133] Next, the system uses the emergency action state transition diagram and the value iteration algorithm of the Markov decision process model to evaluate the value of each emergency path. The system calculates the value evaluation results of each path, including the success probability, time required, resource consumption and potential risks of the path. For example, one path may involve the use of a helicopter for rapid transfer, which has a high probability of success but high resource consumption; another path chooses a ground ambulance, which has a slightly lower probability of success but low resource consumption.

[0134] Finally, based on these value assessment results, the system determines the expected results after each path prediction is executed, and generates multiple decision paths and expected results. For example, the system recommends giving priority to helicopter transfer when weather conditions permit to ensure that the injured can receive professional treatment in the shortest possible time. At the same time, the system also provides backup plans, such as choosing ground ambulances in bad weather, and re-planning resource allocation to ensure the safety and effective treatment of the injured.

[0135] Through this series of operations, the system not only improves the scientificity and accuracy of emergency path assessment, but also enhances the ability to deal with complex situations, greatly improving the success rate and efficiency of treatment.

[0136] Through the above improvements, the present invention significantly improves the scientificity and accuracy of emergency path evaluation. The state space definition and value iteration algorithm of the Markov decision process model ensure a comprehensive evaluation of different paths, covering the success probability, required time, resource consumption and potential risks of the paths. The multiple decision paths and expected results finally generated not only improve the accuracy and flexibility of decision-making, but also enhance the intelligence and coordination level of the emergency system, significantly improving the success rate and efficiency of treatment.

[0137] In order to further improve the scientificity and effectiveness of the first aid strategy, in some embodiments, the 4G / 5G network is used in step 104 to establish a real-time communication connection between the on-site first aid personnel and professional medical personnel, integrate the opinions of multiple experts, improve the optimized first aid strategy, and generate a comprehensive first aid action plan including remote guidance support and on-site first aid action plan, including:

[0138] Utilize 4G / 5G network technology to establish a real-time communication connection between on-site emergency personnel and professional medical personnel to obtain a stable communication link; utilize the stable communication link to transmit the patient's vital signs data, environmental factor assessment results and personalized first aid plans to medical experts to obtain expert reference information from medical experts; integrate the opinions of multiple medical experts in different fields, based on the expert reference information, combine the first aid scenario simulation platform created with virtual reality technology, simulate and adjust the optimized first aid strategy to form a collective wisdom improvement plan; formulate an on-site first aid action plan based on the collective wisdom improvement plan; generate a comprehensive first aid action plan based on the collective wisdom improvement plan and the on-site first aid action plan, including remote guidance support and on-site first aid action plan.

[0139] In this embodiment, 4G / 5G network technology is used to establish a real-time communication connection between on-site emergency personnel and professional medical personnel to ensure high speed, low latency and high reliability of data transmission. A stable communication link refers to a high-quality communication channel established through a 4G / 5G network to ensure that the patient's vital signs data, environmental factor assessment results and personalized first aid plans can be transmitted to medical experts in a timely and accurate manner. Expert reference information is professional advice and guidance provided by medical experts based on the received data, covering multiple aspects such as diagnosis, treatment and care. The first aid scenario simulation platform created by virtual reality technology is used to simulate the actual effects of different first aid measures, helping medical experts and first aid personnel to better understand and adjust first aid strategies. The final comprehensive first aid action plan not only includes remote guidance support, but also details the specific action steps for on-site first aid.

[0140] In the embodiment of the present application, 4G / 5G network technology is first used to establish a real-time communication connection between on-site emergency personnel and professional medical personnel to obtain a stable communication link. Then, through this stable communication link, the patient's vital signs data, environmental factor assessment results and personalized first aid plans are transmitted to medical experts to obtain their expert reference information. Next, the system integrates the opinions of multiple medical experts in different fields, based on these expert reference information, and combines the first aid scenario simulation platform created by virtual reality technology to simulate and adjust the optimized first aid strategy to form a collective wisdom improvement plan. Finally, according to the collective wisdom improvement plan, an on-site first aid action plan is formulated, and a comprehensive first aid action plan including remote guidance support and on-site first aid action plan is generated.

[0141] Here is a specific example:

[0142] In a traffic accident rescue scenario in a remote mountainous area, emergency personnel immediately activated portable environmental monitoring equipment and smart bracelets upon arriving at the scene, and uploaded environmental data such as temperature and humidity and the patient's vital signs to the central control system in real time. The system quickly evaluated the data and found that the temperature at the scene was low and the humidity was high, indicating the possible risk of hypothermia and infection. At the same time, the system initially generated a personalized first aid plan that included measures such as cooling, hemostasis, and infusion.

[0143] Next, the system uses 4G / 5G network technology to establish a real-time communication connection between the first responders on site and the emergency doctors in the hospital, obtaining a stable communication link. Through this link, the first responders transmit the patient's vital signs data, environmental factor assessment results, and personalized first aid plans to medical experts in real time. Emergency doctors and experts in other fields (such as respiratory medicine and surgery) jointly review this data and provide detailed expert reference information, such as adjusting the infusion rate and selecting appropriate hemostatic materials.

[0144] Subsequently, the system integrated the opinions of multiple medical experts in different fields and combined with the emergency scenario simulation platform created by virtual reality technology to simulate and exercise the optimized emergency strategy. For example, the simulation platform demonstrated the effects of different transfer methods and time arrangements, and medical experts proposed further optimization suggestions, such as adding psychological counseling to relieve the tension of the injured. Based on these expert opinions, the system formed a collective wisdom improvement plan and formulated a detailed on-site emergency action plan.

[0145] Finally, the system generated a comprehensive emergency action plan that included remote guidance support and on-site emergency action plan. The plan not only clarified the specific operation methods of each emergency step, but also provided instant guidance and support from remote medical experts to ensure that on-site emergency personnel could make the best decision in the shortest time. Through this series of operations, the emergency plan not only took into account the latest environmental changes and vital signs, but also integrated the professional opinions of multiple experts, greatly improving the success rate and efficiency of treatment.

[0146] Through the above improvements, the present invention significantly improves the scientificity and effectiveness of first aid strategies. The real-time communication connection of 4G / 5G network technology ensures high speed and low latency of data transmission, enabling medical experts to provide professional guidance in a timely manner. The first aid scenario simulation platform created by virtual reality technology enhances the authenticity and accuracy of simulation exercises, helping medical experts and first aid personnel to better understand and adjust first aid strategies. The final generated comprehensive first aid action plan not only includes remote guidance support, but also details the specific action steps for on-site first aid, significantly improving the success rate and efficiency of treatment, and enhancing the intelligence and coordination level of the first aid system.

[0147] This application takes into account that in first aid scenarios, environmental conditions (such as temperature, humidity) and the physiological state of the injured (such as heart rate, blood pressure) have an important impact on the selection of first aid measures. Traditional first aid plans are often based on fixed standard processes, and fail to fully consider individual differences and real-time changing environmental factors, resulting in poor treatment results. In order to improve the speed and accuracy of first aid response, a method is needed that can evaluate environmental and physiological data in real time and generate personalized first aid plans. Therefore, a new optional solution is proposed, which includes:

[0148] The convolutional neural network is used to perform feature extraction and pattern recognition on the first aid input data optimized for the geographical location, and to analyze the optimal combination of first aid measures under different environmental conditions and physiological states, and to generate a preliminary first aid plan, including:

[0149] Using a convolutional neural network to perform feature extraction processing on the geographically optimized emergency input data, and obtaining a multi-layer feature representation through multiple convolutional layer calculations;

[0150] The convolution operation is expressed by the following formula:

[0151]

[0152] Among them, F i (x, y, c) is the feature map output by the i-th convolutional layer, the value of the c-th channel at position (x, y); x, y are the spatial position coordinates; c is the channel index of the output feature map; K is the number of input channels; w, h are the spatial width and height of the convolution kernel; X(x+u, y+v, k) is the value of the k-th channel of the input data X at position (x+u, y+v); K i (u, v, k, c) is the weight of the convolution kernel at position (u, v), from input channel k to output channel c; b i (c) is the bias vector of the i-th convolutional layer; σ is the activation function used to introduce nonlinearity;

[0153] Based on the multi-layer feature representation, a pooling layer is used to reduce computational complexity while retaining important information to generate a simplified feature representation;

[0154] The maximum pooling operation is expressed by the following formula:

[0155]

[0156] Among them, F′(x′, y′, c) is the feature representation after pooling layer processing; x′, y′ are the position coordinates of the feature map after pooling; c is the channel index; S is all the bits in the pooling window; s is the pooling step parameter; F i (x′·s+u, y′·s+v, c) is the value of the corresponding position in the original feature map;

[0157] The simplified feature representation is transformed using a fully connected layer in combination with an activation function to generate a comprehensive feature vector;

[0158] The fully connected layer conversion operation is expressed by the following formula:

[0159]

[0160] Among them, V j is the jth element of the comprehensive feature vector; N is the dimension of the simplified feature representation F′; W fc (i, j) is the fully connected layer weight matrix, representing the weight from input i to output j; F′(i) is the i-th element of the simplified feature representation F′; b fc (j) is the bias term of the fully connected layer;

[0161] According to the comprehensive feature vector, a regression algorithm is applied to analyze the optimal combination of first aid measures under different conditions to obtain a probability distribution of the first aid measures;

[0162] The probability P(m) of selecting the mth combination of first aid measures is calculated by the following formula:

[0163]

[0164] Where P(m) is the probability of selecting the mth combination of first aid measures; M is the number of all first aid measures combinations; θ m is the weight vector corresponding to each combination of first aid measures; V is the comprehensive feature vector; is an exponential function used to calculate the relative probability of each combination of first aid measures;

[0165] Based on the probability distribution, several combinations of first aid measures with the highest probability are selected, and after considering the practical feasibility, a preliminary first aid plan is generated.

[0166] The following is a detailed explanation of each parameter:

[0167] F i (x, y, c) is the feature map output by the i-th convolutional layer, the value of the c-th channel at position (x, y). This is the local feature representation calculated by the convolution operation. It is calculated pixel by pixel by convolution operation.

[0168] σ introduces nonlinearity, which enables the model to capture more complex patterns. Activation functions such as ReLU, Sigmoid, etc. can break the linear relationship and increase the model's expressiveness.

[0169] It is the sum of all input channels. Each convolution kernel extracts information from multiple input channels to ensure the fusion of multi-dimensional features.

[0170] It is the spatial range of the convolution kernel. The convolution kernel slides in the spatial range to capture local features.

[0171] X(x+u, y+v, k) is the value of the kth channel of the input data X at position (x+u, y+v). It is directly obtained from the input data.

[0172] K i (u, v, k, c) is the weight of the convolution kernel at position (u, v), from input channel k to output channel c. It is learned through the training process.

[0173] b i (c) is the bias vector of the i-th convolutional layer. It is learned through the training process. The bias term is used to adjust the threshold of the activation function and enhance the flexibility of the model.

[0174]

[0175] F′(x′, y′, c) is the feature representation after the pooling layer. It is obtained by applying the maximum pooling operation to the original feature map.

[0176] x′, y′ are the position coordinates of the feature map after pooling. Reducing the spatial resolution reduces the computational complexity.

[0177] c is the channel index. It holds multi-channel information.

[0178] S is all the bits within the pooling window. It defines the area range of the pooling operation.

[0179] s is the pooling step size parameter, which controls the distance the pooling window moves and affects the size of the output feature map.

[0180] F i (x′·s+u, y′·s+v, c) is the value of the corresponding position in the original feature map. It is directly obtained from the original feature map.

[0181]

[0182] V j is the jth element of the comprehensive feature vector. It is obtained by transforming through the fully connected layer.

[0183] N is the dimension of the simplified feature representation F′. It represents the number of input features.

[0184] W fc (i, j) is the fully connected layer weight matrix, which represents the weight from input i to output j. It is learned through the training process.

[0185] F′(i) is the i-th element of the simplified feature representation F′. It is directly obtained from the output of the pooling layer.

[0186] b fc (j) is the bias term of the fully connected layer. It is learned through the training process. It is similar to the bias term in the convolutional layer and is used to adjust the threshold of the activation function.

[0187]

[0188] P(m) is the probability of selecting the mth combination of emergency measures. The comprehensive feature vector is mapped to a probability distribution through the softmax function, which makes it easier to select the optimal solution.

[0189] M is the number of all first aid measures combinations. It represents the number of possible decision paths.

[0190] θ m is the weight vector corresponding to each combination of first aid measures. It is learned through the training process. The weight vector reflects the importance of different first aid measures.

[0191] V is the comprehensive feature vector. It is obtained by transformation through the fully connected layer. The comprehensive feature vector contains all the important features and is used to evaluate the effectiveness of different first aid measures.

[0192] is an exponential function used to calculate the relative probability of each combination of first aid measures. The exponential function ensures that the probability value is positive and is normalized to a probability distribution through the softmax function.

[0193] Here is a specific example:

[0194] In a traffic accident rescue scenario in the city center, emergency personnel immediately activated portable environmental monitoring equipment and smart bracelets upon arriving at the scene, and uploaded environmental data such as temperature and humidity and the vital signs of the injured to the central control system in real time. The system quickly evaluated the data and found that the temperature and humidity at the scene were high, indicating a possible risk of heat stroke. At the same time, the system initially generated a personalized first aid plan that included measures such as cooling, hemostasis, and infusion.

[0195] Assume we have the following initial data:

[0196] Environmental data: temperature T = 35 ° C, humidity H = 80%

[0197] Vital signs data: heart rate HR = 120 beats / minute, blood pressure BP = 140 / 90 mmHg

[0198] Geographical location information: 5 kilometers from the nearest hospital

[0199] First, a convolutional neural network is used to extract features from these data. Assume that the dimension of the input data X is 64×64×3, where 64×64 is the spatial resolution and 3 is the number of channels (corresponding to temperature, humidity, and vital signs, respectively). The convolution kernel size is 3×3, the step size is 1, the padding is 0, and the number of output channels is 16.

[0200] According to the convolution operation formula:

[0201]

[0202] Calculate the feature map F output by the i-th convolutional layer i :

[0203] For example, the value at position (32, 32) and channel c=5 is:

[0204]

[0205] Next, a max pooling layer is used to reduce computational complexity while retaining important information;

[0206] Assume that the pooling window size is 2×2 and the stride is 2;

[0207] According to the maximum pooling operation formula:

[0208]

[0209] For example, the value at position (16, 16) and channel c=5 is:

[0210]

[0211] Then, the fully connected layer is used in combination with the activation function to transform the simplified feature representation to generate a comprehensive feature vector. Assume that the dimension of the simplified feature representation F′ is 1024, and the fully connected layer weight matrix W fc The dimension is 1024×128, and the bias term b fc The dimension is 128;

[0212] According to the fully connected layer conversion operation formula:

[0213]

[0214] For example, the 64th element of the comprehensive feature vector V is:

[0215]

[0216] Finally, according to the comprehensive feature vector V, the regression algorithm is used to analyze the optimal combination of first aid measures under different conditions and obtain the probability distribution of first aid measures;

[0217] Assume there are 5 combinations of emergency measures M, and the weight vector θ of each combination is m The dimension is 128;

[0218] According to the probability formula of selecting the mth combination of first aid measures:

[0219]

[0220] For example, the probability of choosing the third combination of first aid measures is:

[0221]

[0222] Assume that the calculation results show:

[0223] P(1)=0.1, P(2)=0.2, P(3)=0.4, P(4)=0.15, P(5)=0.15

[0224] Based on these probability distributions, the system selects several combinations of first aid measures with the highest probability, and generates a preliminary first aid plan after considering the actual feasibility of the operation. For example, the system recommends taking the third combination of first aid measures first, because it has the highest selection probability (0.4), and this combination includes cooling, hemostasis, and infusion measures, which meet the needs of the current environment and physiological state.

[0225] Through this series of operations, the first aid plan not only takes into account the latest environmental changes and vital signs, but also integrates the professional opinions of multiple experts, greatly improving the success rate and efficiency of treatment.

[0226] This application takes into account that in first aid scenarios, rapid changes in environmental conditions (such as sudden weather changes, traffic jams) and the development of the patient's condition have a direct impact on the success rate and safety of rescue operations. Traditional first aid plans are often based on fixed processes and are difficult to adapt to dynamically changing environments. In order to improve the flexibility and scientificity of decision-making, a method is needed that can adjust first aid measures based on real-time environmental predictions. Therefore, a new optional solution is proposed, which includes:

[0227] According to the environmental change prediction results, combined with the combination of measures in the personalized first aid plan, the Markov decision process model is used for analysis and processing to evaluate the success probability and potential risks of each path, and generate multiple decision paths and expected results, including:

[0228] Integrate the environmental change prediction results and the measures in the personalized first aid plan to obtain comprehensive first aid decision input data;

[0229] Define a state space based on the comprehensive emergency decision input data, including all emergency actions as state transitions;

[0230] The state transition probability T is expressed by the following formula: MDP (s, a, s′):

[0231] T MDP (s, a, s′) = P (s′∣s, a)

[0232] Among them, P(s′|s,a) is the probability of transitioning from state s to state s′ after taking action a;

[0233] The immediate reward R obtained by taking action a in state s is calculated by the following formula: MDP (s, a):

[0234]

[0235] Among them, w s , w t , w c , w r are weight parameters, which respectively represent the importance of success probability, required time, resource consumption and potential risks; P success (a) represents the success probability of action a; λ represents the time decay coefficient, which is used to adjust the influence of time; T a Indicates the time required for action a; C a represents the resource consumption of action a; R a Indicates the potential risk of action a;

[0236] The value assessment result V is calculated by the following formula: MDP (s):

[0237]

[0238] Among them, V MDP (s) is the value assessment result; γ MDP is the discount factor, indicating the degree of decay of future rewards (0<γ MDP <1);

[0239] According to the value assessment results V MDP (s), state transition probability T MDP (s, a, s′) and the immediate reward R MDP (s, a), screen out several decision paths with high success probability and low potential risk, and predict the expected result after execution for each path, and finally generate multiple decision paths and expected results;

[0240] The probability of choosing the pth decision path is calculated using the following formula:

[0241]

[0242] Among them, Ppath (p) is the probability of choosing the pth decision path; Q is the number of all decision paths; θ p The weight vector corresponding to each decision path; V MDP is the state value function.

[0243] The following is a detailed explanation of each parameter:

[0244] State transition probability formula

[0245] T MDP (s, a, s′) = P (s′∣s, a)

[0246] T MDP (s, a, s') is the probability of transitioning from state s to state s' after taking action a. It is obtained through historical data statistics or simulation.

[0247] P(s′|s, a) is the state transition probability, that is, the probability of reaching the new state s′ after executing action a in state s. It is used to define the state transition mechanism in the Markov decision process and is the basis of the decision model.

[0248]

[0249] R MDP (s, a) is the immediate reward obtained by taking action a in state s. It is calculated according to the above formula.

[0250] w s , w t , w c , w r are weight parameters that represent the importance of success probability, time required, resource consumption, and potential risks. They are determined through expert evaluation or training process adjustment. They are used to balance the impact of different factors on the total reward and ensure that the decision takes multiple factors into consideration.

[0251] P success (a) is the success probability of action a. It is based on historical success rate statistics or prediction models. The success probability directly affects the possibility of task completion, so it is an important evaluation indicator.

[0252] λ is the time decay coefficient, which is used to adjust the impact of time. It can be set according to the actual situation or optimized through experiments. It reflects the urgency of time. The sooner the task is completed, the better.

[0253] T a is the time required for action a. Based on historical records or estimates. Time cost is one of the important factors to be considered in emergency operations.

[0254] C ais the resource consumption of action a. Based on resource usage statistics, the efficient use of resources is crucial for emergency resource management.

[0255] R a It is the potential risk of action a. Based on risk assessment models or historical data. Risk control is an important part of any action.

[0256]

[0257] V MDP (s) is the value evaluation result, which is obtained by iterative algorithms such as value iteration or policy iteration.

[0258] γ MDP is the discount factor, indicating the degree of decay of future rewards (0<γ MDP <1). Set according to the application scenario. The value of future rewards is usually considered lower than current rewards, and the discount factor is used to quantify this difference.

[0259]

[0260] P path (p) is the probability of choosing the pth decision path. It is calculated using the above formula.

[0261] Q is the number of all decision paths. It is used for normalization so that the sum of the selection probabilities of all paths is 1.

[0262] θ p The weight vector corresponding to each decision path is determined by a learning or optimization process. The weight vector reflects the importance or preference of different paths.

[0263] V MDP It is the state value function. It is calculated by the value evaluation result. It is used as a criterion to measure the pros and cons of different paths.

[0264] The following is an introduction to the design reasons of each sub-item:

[0265] w s ·log(P success (a)): The logarithmic function is used to avoid too large a reward value when the success probability is close to 1, while ensuring a large penalty when the success probability is low. This helps to balance the differences between different success probabilities. s To adjust the proportion of success probability in the total reward to make it more in line with actual needs.

[0266] Exponential decay function It is used to quantify the impact of time. The shorter the time, the higher the reward, and vice versa. The time decay coefficient λ controls the degree of time impact.t To adjust the proportion of time efficiency in the total reward, so that decisions pay more attention to timeliness.

[0267] w c ·C a : A direct linear relationship means that resource consumption is inversely proportional to the reward, that is, the less resource consumption, the higher the reward. c To adjust the proportion of resource consumption in the total reward and encourage the efficient use of resources.

[0268] w r ·R a : A direct linear relationship means that the potential risk is inversely proportional to the reward, that is, the smaller the risk, the higher the reward. r To adjust the proportion of risk in the total reward and emphasize the importance of risk control.

[0269] By adding up the sub-items, multiple factors are taken into consideration. Each sub-item represents a different dimension of consideration (probability of success, time, resource consumption, risk). Adding them together to obtain the total immediate reward can comprehensively evaluate the pros and cons of an action. By setting different weight parameters, the relative importance of each factor in the decision-making can be flexibly adjusted to adapt to different application scenarios and needs.

[0270] The overall design of the formula aims to establish a comprehensive evaluation system to quantify the immediate effect of each combination of first aid measures. By introducing multiple dimensions (probability of success, time, resource consumption, risk) and assigning corresponding weights, the formula can not only accurately reflect the impact of each factor on decision-making, but also adapt to different emergency situations and environmental changes by adjusting the weights. The ultimate goal is to generate scientific and reasonable first aid strategies, improve the success rate and efficiency of rescue, and minimize resource waste and risks.

[0271] In this way, the system can quickly evaluate multiple possible courses of action in complex emergency scenarios and select the optimal path, thereby improving the effectiveness and speed of treatment.

[0272] Here is a specific example:

[0273] In a traffic accident rescue scenario in a city center, emergency personnel immediately start the portable environmental monitoring equipment after arriving at the scene, and upload environmental data such as temperature and humidity to the central control system in real time. The system quickly evaluates this data and predicts the trend of environmental changes in the future. At the same time, it combines personalized emergency plans and uses the Markov decision process model for analysis and processing, and finally generates multiple decision paths and their expected results.

[0274] Assume we have the following initial data:

[0275] Environmental change forecast: Temperature will drop by 5°C and humidity will increase by 10% in the next hour

[0276] Personalized first aid plan: including cooling, hemostasis, infusion and other measures

[0277] Current status: The temperature and humidity at the scene are high, indicating a possible risk of heat stroke; the victim's heart rate is 120 beats / minute and blood pressure is 140 / 90 mmHg

[0278] Target state s′: The patient's vital signs are stable and body temperature returns to normal

[0279] First, define the state space, which includes all emergency actions as state transitions. For example, from the current state s to the target state s′, the state transition probability T of taking cooling measure a is MDP (s, a, s′) is given by:

[0280] T MDP (s, a, s′) = P (s′∣s, a)

[0281] Assume that after data analysis, the probability of successfully transferring to the target state after taking cooling measures is P(s′|s, a)=0.8;

[0282] Then, calculate the immediate reward R obtained by taking action a in state s MDP (s, a). Assume that the weight parameters are w s =0.6, w t =0.2, w c =0.1, w r = 0.1, time decay coefficient λ = 0.1, success probability of action a P success (a) = 0.9, time required T a = 10 minutes, resource consumption C a =5, potential risk R a =0.3.

[0283] Instant Rewards MDP (s, a) is given by:

[0284]

[0285] R MDP (s, a)=0.6·log(0.9)+0.2·e -0.1×10 +0.1·5+0.1·0.3

[0286] Substitute the values ​​to calculate:

[0287] R MDP(s, a)≈0.6·(-0.10536)+0.2·0.36788+0.5+0.03

[0288] R MDP (s, a)≈-0.0632+0.0736+0.5+0.03

[0289] R MDP (s, a)≈0.5404

[0290] Next, calculate the value assessment result V MDP (s). Assume the discount factor γ MDP =0.9, and the value evaluation results of all possible states s′ have been calculated;

[0291] Value Assessment Results MDP (s) is given by:

[0292]

[0293] Assume that after iterative calculation, V MDP (s)≈0.6;

[0294] Finally, several decision paths with high success probability and low potential risk are screened out, and the expected results after execution are predicted for each path. Assume that there are 3 decision paths in total, and their value assessment results are V MDP，1 =0.6, V MDP，2 =0.55, V MDP，3 =0.5.

[0295] The probability of choosing the pth decision path is P path (p) is given by:

[0296]

[0297] Assume that the weight vector θ of all paths p =1, then:

[0298]

[0299] P path (1)≈0.35

[0300] Similarly, the selection probabilities of the other two paths can be calculated.

[0301] Assume that the calculation results show:

[0302] P path (1)≈0.35, P path (2)≈0.33, P path (3)≈0.32

[0303] Based on these probability distributions, the system selects the decision path 1 with the highest probability and predicts the expected results after its execution. For example, this path recommends taking cooling measures first because it has a higher selection probability (about 0.35) and the path has a high probability of success and low potential risk, which meets the needs of the current environment and physiological state.

[0304] Through this series of operations, the first aid plan not only takes into account the latest environmental changes and vital signs, but also integrates the professional opinions of multiple experts, greatly improving the success rate and efficiency of treatment.

[0305] Figure 2 A schematic diagram of the structure of an emergency rescue system based on environmental adaptability is provided for an embodiment of the present application. Figure 2 As shown, the system includes:

[0306] The collection module 21 is used to collect environmental information of temperature and humidity at the accident site, evaluate the additional risk caused by the environment to the injured, and obtain the environmental factor evaluation result;

[0307] An analysis module 22 is used to generate a personalized first aid plan by using a deep learning algorithm to analyze and process the environmental factor assessment results and the patient's vital sign data obtained from the wearable medical device in combination with the nearest medical institution location information provided by the geolocation service;

[0308] The prediction module 23 is used to continuously monitor and instantly analyze changes in environmental conditions based on a dynamic programming algorithm, predict the results under different decision paths, conduct a rehearsal test on the first aid scenario simulation platform created by virtual reality technology for the personalized first aid plan, and optimize and adjust the first aid strategy according to the rehearsal results to obtain an optimized first aid strategy;

[0309] The integration module 24 is used to establish a real-time communication connection between on-site emergency personnel and professional medical personnel using the 4G / 5G network, integrate the opinions of multiple experts to form collective wisdom, further improve the optimized emergency strategy, and generate a comprehensive emergency action plan that includes remote guidance support and on-site emergency action plan.

[0310] Figure 2 The first aid system based on environmental adaptability can be implemented Figure 1 The implementation principle and technical effect of the first aid method based on environmental adaptability described in the illustrated embodiment will not be described in detail. The specific manner in which each module and unit performs operations in the first aid system based on environmental adaptability in the above embodiment has been described in detail in the embodiment of the method, and will not be described in detail here.

[0311] In one possible design, Figure 2The emergency rescue system based on environmental adaptability of the embodiment shown can be implemented as a computing device, such as Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32;

[0312] The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 32 .

[0313] The processing component 32 is used for the above Figure 1 The first aid method based on environmental adaptability in the embodiment described above.

[0314] The processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components to perform the above method.

[0315] The storage component 31 is configured to store various types of data to support operations at the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0316] Of course, the computing device may also include other components, such as input / output interfaces, display components, communication components, etc.

[0317] The input / output interface provides an interface between the processing component and the peripheral interface module, which may be an output device, an input device, etc.

[0318] The communication component is configured to facilitate, among other things, wired or wireless communications between the computing device and other devices.

[0319] Among them, the computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. In this case, the computing device can refer to a cloud server, and the above-mentioned processing components, storage components, etc. can be basic server resources rented or purchased from the cloud computing platform.

[0320] The present application also provides a computer storage medium storing a computer program, wherein the computer program can achieve the above-mentioned Figure 1The first aid method based on environmental adaptability of the illustrated embodiment.

[0321] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0322] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0323] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0324] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A first aid method based on environmental adaptability, characterized in that: include: Collect environmental information on temperature and humidity at the accident site, assess the additional risks posed by the environment to the injured, and obtain environmental factor assessment results; Based on the environmental factor assessment results and the patient's vital sign data obtained from the wearable medical device, combined with the location information of the nearest medical institution provided by the geo-positioning service, a deep learning algorithm is used for analysis and processing to generate a personalized first aid plan; Based on the dynamic programming algorithm, the changes in environmental conditions are continuously monitored and analyzed in real time, the results of the personalized first aid plan under different decision paths are predicted, and the first aid scenario simulation platform created by virtual reality technology is rehearsed for the personalized first aid plan. The first aid strategy is optimized and adjusted according to the rehearsal results to obtain the optimized first aid strategy; Use 4G / 5G networks to establish real-time communication connections between on-site emergency personnel and professional medical personnel, integrate opinions from multiple experts, improve the optimized emergency strategy, and generate a comprehensive emergency action plan that includes remote guidance support and on-site emergency action plans.

2. The method according to claim 1, characterized in that The environmental factor assessment results and the patient's vital signs data obtained from the wearable medical device are combined with the nearest medical institution location information provided by the geo-location service, and a deep learning algorithm is used for analysis and processing to generate a personalized first aid plan, including: Integrate the environmental factor assessment results with the patient's vital sign data obtained from the wearable medical device to obtain comprehensive first aid input data; Based on the nearest medical institution location information provided by the geo-location service, the comprehensive emergency input data is associated and processed, and the geographical location-optimized emergency input data is generated in combination with the influence of the geographical location on the emergency measures; Using a convolutional neural network to perform feature extraction and pattern recognition on the geographically optimized first aid input data, and analyzing the optimal combination of first aid measures under different environmental conditions and physiological states to generate a preliminary first aid plan; The preliminary first aid plan is matched and adjusted based on reasoning technology of historical first aid cases to generate a personalized first aid plan.

3. The method according to claim 2, characterized in that The convolutional neural network is used to extract features and recognize patterns of the geographically optimized emergency input data, and to analyze the optimal combination of emergency measures under different environmental conditions and physiological states to generate a preliminary emergency plan, including: Using a convolutional neural network to perform feature extraction processing on the geographically optimized emergency input data to obtain a multi-layer feature representation; Using a pooling layer to simplify the multi-layer feature representation to obtain a simplified feature representation; The simplified feature representation is transformed using a fully connected layer in combination with an activation function to generate a comprehensive feature vector; According to the comprehensive feature vector, a regression algorithm is applied to analyze the optimal combination of first aid measures under different conditions to obtain the probability distribution of first aid measures, and based on the probability distribution, a preliminary first aid plan is generated.

4. The method according to claim 1, characterized in that: The dynamic programming algorithm is used to continuously monitor and instantly analyze changes in environmental conditions, predict the results of the personalized first aid plan under different decision paths, conduct a rehearsal test on the first aid scenario simulation platform created by virtual reality technology for the personalized first aid plan, and optimize and adjust the first aid strategy according to the rehearsal results to obtain an optimized first aid strategy, including: Using a dynamic programming algorithm, the real-time environmental condition changes at the accident site are continuously monitored and analyzed in real time, and the environmental challenges faced by the personalized first aid plan at different time points and under different conditions are predicted to obtain environmental change prediction results; According to the environmental change prediction results, combined with the combination of measures in the personalized first aid plan, the Markov decision process model is used for analysis and processing to evaluate the success probability and potential risks of each path, and generate multiple decision paths and expected results; A first aid scenario simulation platform created based on virtual reality technology is used to perform rehearsal tests on the multiple decision paths and expected results, verify the actual effects of different decision paths, and obtain rehearsal test feedback information; According to the feedback information from the rehearsal test, the personalized first aid plan is optimized and adjusted in a targeted manner to generate an optimized first aid strategy, wherein the adjustment content includes the adjustment of the sequence of first aid steps and the re-planning of resource allocation.

5. The method according to claim 4, characterized in that The dynamic programming algorithm is used to continuously monitor and analyze the real-time environmental condition changes at the accident site, predict the environmental challenges faced by the personalized first aid plan at different time points and under different conditions, and obtain environmental change prediction results, including: Using a sensor network deployed on-site or on a remote server to collect real-time environmental parameters at the accident site, wherein the real-time environmental parameters include temperature, humidity and air quality; Based on the real-time environmental parameters, a state model of the environmental conditions at the accident site is constructed using the state space representation method in the dynamic programming algorithm, wherein the state model is used to describe the changing rules of the environmental conditions over time and space, and to generate an environmental state model; Using the value iteration method in the dynamic programming algorithm and the environmental state model, the changing trend of environmental conditions is predicted, and the environmental challenges faced by the personalized first aid program at different time points and under different conditions are predicted, so as to obtain the time series prediction results of environmental changes; Based on the time series prediction results of the environmental changes, key environmental factors and change trends that have a significant impact on emergency operations are identified, and environmental change prediction results are generated.

6. The method according to claim 5, characterized in that The method uses the Markov decision process model to analyze and process the predicted environmental change results in combination with the combination of measures in the personalized first aid plan, evaluates the success probability and potential risks of each path, and generates multiple decision paths and expected results, including: Integrate the environmental change prediction result and the measure combination in the personalized first aid plan to obtain comprehensive first aid decision input data; Based on the Markov decision process model, the state space definition processing is performed on the comprehensive emergency decision input data, all emergency actions are identified as state transitions, and the probability distribution of taking different actions under each state is determined to generate an emergency action state transition diagram; By using the emergency action state transition diagram and combining the value iteration algorithm of the Markov decision process model, each emergency path is evaluated for value, and the value evaluation results of each path are generated, wherein the value evaluation content includes the success probability, required time, resource consumption and potential risks of the path; According to the value assessment results of each path, the expected result after each path prediction execution is determined, and multiple decision paths and expected results are generated.

7. The method according to claim 1, characterized in that The 4G / 5G network is used to establish a real-time communication connection between on-site emergency personnel and professional medical personnel, integrate the opinions of multiple experts, improve the optimized emergency strategy, and generate a comprehensive emergency action plan including remote guidance support and on-site emergency action plan, including: Using 4G / 5G network technology, establish real-time communication connection between on-site emergency personnel and professional medical personnel to obtain a stable communication link; Utilizing the stable communication link, the patient's vital sign data, environmental factor assessment results, and personalized first aid plans are transmitted to medical experts to obtain expert reference information from the medical experts; Integrate the opinions of multiple medical experts in different fields, based on the expert reference information, and combine the emergency scenario simulation platform created with virtual reality technology to simulate and adjust the optimized emergency strategy to form a collective wisdom improvement plan; Formulate an on-site emergency action plan based on the collective wisdom improvement plan; Based on the collective wisdom improvement plan and the on-site first aid action plan, a comprehensive first aid action plan including remote guidance support and on-site first aid action plan is generated.

8. An environmental adaptability-based first aid system, characterized in that: include: The collection module is used to collect environmental information such as temperature and humidity at the accident site, evaluate the additional risk of the environment to the injured, and obtain the environmental factor evaluation results; An analysis module, for generating a personalized first aid plan by using a deep learning algorithm to analyze and process the environmental factor assessment results and the patient's vital sign data obtained from the wearable medical device in combination with the nearest medical institution location information provided by the geolocation service; A prediction module is used to continuously monitor and instantly analyze changes in environmental conditions based on a dynamic programming algorithm, predict the results under different decision paths, conduct a rehearsal test on the first aid scenario simulation platform created by virtual reality technology for the personalized first aid plan, and optimize and adjust the first aid strategy according to the rehearsal results to obtain an optimized first aid strategy; The integration module is used to establish a real-time communication connection between on-site emergency personnel and professional medical personnel using the 4G / 5G network, integrate the opinions of multiple experts to form collective wisdom, further improve the optimized emergency strategy, and generate a comprehensive emergency action plan that includes remote guidance support and on-site emergency action plan.

9. A computing device, characterized in that It comprises a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement a first aid method based on environmental adaptability as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a computer, the first aid method based on environmental adaptability as claimed in any one of claims 1 to 7 is implemented.

Images