First aid method and system based on 5G communication and edge device

Abstract

The application provides a first-aid method and system based on 5G communication and an edge device, and the method comprises the following steps: determining a communication protocol target adaptation protocol with a 5G base station based on the edge device; acquiring multi-modal data according to the target adaptation protocol based on the edge device, and performing priority classification on the multi-modal data to obtain a priority classification result; evaluating 5G network resources of the edge device to determine a resource evaluation result, and determining a cloud transmission strategy based on the resource evaluation result and the priority classification result; performing correlation analysis on vital sign data, medical image data and voice recording data to obtain a computing task and a task complexity, and processing the computing task based on the task complexity and the cloud transmission strategy to determine a data processing result; and transmitting the data processing result to a first-aid intelligent center and a mobile terminal in a hierarchical manner. The application solves the compatibility problem existing between non-uniform device interfaces, and improves the processing capacity of the edge device.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of communication technology, in particular to an emergency method and system based on 5G communication and edge device. BACKGROUND

[0002] With the rapid development and wide application of 5G communication technology, in the application field of mobile communication technology, 5G communication technology brings new possibilities for data transmission and interaction in emergency scenarios with its high speed, low latency and large capacity characteristics; edge devices effectively reduce data transmission pressure and response time by realizing fast processing at the data source. The application of 5G communication and edge device in emergency system can realize real-time transmission of key information such as patient vital signs and on-site video, assist remote diagnosis and emergency guidance, and greatly improve the efficiency of emergency.

[0003] In practical application, edge device needs to cooperate closely with 5G communication network when preprocessing emergency data. However, the interface and protocol between edge device and 5G network have not been completely unified, which may cause compatibility problems between devices. For example, edge devices produced by different manufacturers may not be able to be connected smoothly with 5G base stations, affecting the transmission and processing efficiency of data. In addition, edge device may have insufficient processing capacity when processing a large amount of emergency data, especially in the analysis of complex medical data, which may affect subsequent treatment decisions and thus the timeliness of emergency. SUMMARY

[0004] The present application provides an emergency method and system based on 5G communication and edge device, which solves the compatibility problem between device interfaces and improves the processing capacity of edge device.

[0005] In the first aspect, the present application provides an emergency method based on 5G communication and edge device, comprising:

[0006] detecting the communication protocol type of 5G base station based on edge device, and adapting the protocol of edge device and 5G base station based on the detection result to determine the target adaptation protocol;

[0007] acquiring multi-modal data based on the target adaptation protocol by edge device, and classifying the multi-modal data by priority to obtain priority classification result; the multi-modal data includes vital sign data, medical image data and voice recording data;

[0008] evaluating the 5G network resources of edge device to determine resource evaluation result, and determining cloud transmission strategy based on the resource evaluation result and the priority classification result;

[0009] correlatively analyze the vital sign data, the medical image data and the voice recording data to obtain a computing task and a task complexity, and process the computing task based on the task complexity and the cloud transmission strategy to determine a data processing result;

[0010] transmit the data processing result to the emergency intelligent center and the mobile terminal in a hierarchical manner.

[0011] In a second aspect, the present application further provides an emergency method system based on 5G communication and edge equipment, which is applied to the emergency method based on 5G communication and edge equipment as described in the first aspect. The emergency method system based on 5G communication and edge equipment comprises:

[0012] a protocol adaptation module, configured to detect a communication protocol type of a 5G base station based on an edge equipment, and to adapt a protocol of the edge equipment and the 5G base station based on a detection result to determine a target adaptation protocol;

[0013] a priority classification module, configured to obtain multi-modal data based on the edge equipment according to the target adaptation protocol, and to classify the multi-modal data in priority to obtain a priority classification result; the multi-modal data comprises vital sign data, medical image data and voice recording data;

[0014] a strategy determination module, configured to evaluate 5G network resources of the edge equipment to determine a resource evaluation result, and to determine a cloud transmission strategy based on the resource evaluation result and the priority classification result;

[0015] a data processing module, configured to correlatively analyze the vital sign data, the medical image data and the voice recording data to obtain a computing task and a task complexity, and to process the computing task based on the task complexity and the cloud transmission strategy to determine a data processing result;

[0016] a data delivery module, configured to transmit the data processing result to the emergency intelligent center and the mobile terminal in a hierarchical manner.

[0017] In a third aspect, the present application further provides an electronic device, comprising: a memory configured to store a computer software program; and a processor configured to read and execute the computer software program to realize the emergency method based on 5G communication and edge equipment as described in any one of the above aspects.

[0018] In a fourth aspect, the present application further provides a non-transitory computer readable storage medium, wherein the storage medium stores a computer software program, and the computer software program is executed by a processor to realize the emergency method based on 5G communication and edge equipment as described in any one of the above aspects.

[0019] In a fifth aspect, the present application also provides a computer program product comprising a computer program which, when executed by a processor, implements the method for emergency treatment based on 5G communication and edge device according to any one of the above.

[0020] The method for emergency treatment based on 5G communication and edge device provided by the embodiments of the present application ensures the efficiency of data transmission by protocol adaptation between the edge device and the 5G base station to obtain a target adaptation protocol corresponding to the edge device, avoids compatibility problems between device interfaces that are not unified, further reasonably allocates computing tasks according to the transmission priority of multi-modal data, the evaluation result of the 5G network and the task complexity handled by the edge device, realizes local processing of simple tasks to reduce delay, cloud processing of complex tasks to ensure the accuracy of data transmission and processing, thereby improving the processing capability of the edge device, avoiding the situation that the processing capability of the edge device is insufficient when analyzing data of complex conditions, and further ensuring that the emergency decision is not affected, ensuring the timeliness of emergency treatment and improving the efficiency of emergency treatment. BRIEF DESCRIPTION OF DRAWINGS

[0021] Figure 1 is a flowchart of the method for emergency treatment based on 5G communication and edge device provided by the embodiments of the present application;

[0022] Figure 2 is a structural schematic diagram of the system for emergency treatment based on 5G communication and edge device provided by the embodiments of the present application;

[0023] Figure 3 is an embodiment diagram of an electronic device provided by the embodiments of the present application;

[0024] Figure 4 is an embodiment diagram of a computer readable storage medium provided by the embodiments of the present application. DETAILED DESCRIPTION

[0025] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0026] In the description of the present application, the terms "first", "second" are used only for the purpose of description, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more features. In the description of the present application, the meaning of "multiple" is two or more, unless otherwise specifically limited.

[0027] In the description of the present application, the term "for example" is used to indicate "as an example, instance, or illustration." Any embodiment described as "for example" in this application is not necessarily to be construed as preferred or advantageous over other embodiments. The following description is presented to enable any person skilled in the art to make and use the present application. In the following description, for the purpose of explanation, details are set forth. It is apparent to those skilled in the art that the present application can be practiced without the use of these specific details. In other instances, well-known structures and processes are not elaborated in order not to obscure the description of the present application with unnecessary details. Therefore, the present application is not intended to be limited to the embodiments shown, but is to be accorded the widest scope consistent with the principles and features disclosed.

[0028] Reference Figure 1 , Figure 1 is a flowchart of the emergency method based on 5G communication and edge device provided by the present application. The execution subject of the emergency method based on 5G communication and edge device in the embodiment of the present application is an emergency system. Therefore, the emergency method based on 5G communication and edge device comprises:

[0029] Step 10, detecting the communication protocol type of the 5G base station based on the edge device, and protocol adapting the edge device and the 5G base station based on the detection result to determine the target adaptation protocol.

[0030] Optionally, since the 5G base station contains multiple protocol types, such as non-standalone networking (NSA) and standalone networking (SA), different protocols have differences in bandwidth, delay, reliability, etc. Therefore, after the emergency system is started, the emergency system first detects the communication protocol type of the 5G base station based on the edge device. Specifically, the edge device sends a probe signal to the 5G base station and analyzes the signal returned by the 5G base station, so as to match the analysis result with the pre-stored protocol, thereby determining the target adaptation protocol. Specifically, as described in steps 101-105. Through protocol adaptation, an efficient and stable communication link between the edge device and the 5G base station is established, the protocol compatibility problem between devices of different manufacturers is solved, and the problem of data transmission interruption or low efficiency caused by protocol mismatch is avoided, thereby laying a solid foundation for subsequent transmission of emergency data.

[0031] Further, in an embodiment, in a certain emergency scene, the edge device on the emergency vehicle or emergency device sends a probe signal packet to the nearby 5G base station, the base station returns a signal, and the edge device analyzes the returned signal and matches it with the protocol in the protocol library built in the system, and finally determines to adjust the signal modulation mode to the orthogonal frequency division multiplexing (OFDM) mode suitable for the TDD protocol, completes the protocol adaptation, and successfully establishes a high-speed stable connection.

[0032] Step 20, based on the edge device, the target adaptation protocol is used to obtain multi-modal data, and the multi-modal data is prioritized to obtain a priority classification result; the multi-modal data includes vital sign data, medical image data and voice recording data.

[0033] Optionally, after determining the target adaptation protocol, the edge device in the emergency system obtains multi-modal data through various sensors and acquisition devices, such as obtaining vital sign data (heart rate, blood pressure, blood oxygen, etc.) through an electrocardiogram sensor, obtaining medical image data through a portable ultrasound device, and obtaining voice recording data (dialogue between medical staff and patients, on-site situation description, etc.) through a microphone. After obtaining the data, the data is prioritized according to the urgency of the data and the timeliness requirement of the emergency decision, as described in steps 201-204. Prioritizing multi-modal data can prioritize the transmission of urgent and critical data in the case of limited 5G network resources.

[0034] Step 30, the 5G network resources of the edge device are evaluated, the resource evaluation result is determined, and the cloud transmission strategy is determined based on the resource evaluation result and the priority classification result.

[0035] Optionally, after the edge device in the emergency system starts collecting multi-modal data of the patient, it starts uploading the data. At this time, the edge device starts to evaluate the 5G network resources in this area before uploading the data, including network bandwidth, delay, packet loss rate and node load, etc. Resource index data, such as larger bandwidth, lower delay, smaller packet loss rate, indicating more abundant network resources, etc. Evaluation results, as described in steps 301-305.

[0036] Further, after the emergency system obtains the resource evaluation result, the priority classification result obtained in the previous step is combined to develop a cloud transmission strategy. For example, if the network resources are sufficient, all data can be transmitted to the cloud in order of priority; if the network resources are tight, the high-priority vital sign data is transmitted first, and the medium and low-priority data is compressed, buffered or delayed. For example, the medical image data is compressed using a compression algorithm before transmission, and the voice recording data is temporarily buffered in the edge device and sent when the network resources are idle. The data transmission strategy is dynamically adjusted according to the actual network situation, making full use of network resources, ensuring timely transmission of important data, avoiding data loss or delay due to network congestion, and improving the efficiency and stability of data transmission.

[0037] Further, in an embodiment, taking the scene of first aid in an outdoor large-scale event as an example, due to a large number of devices simultaneously accessing the 5G network, network resources are strained. Therefore, the edge device assesses that the network bandwidth has decreased to 50% of the normal level, and the delay has increased. At this time, the edge device transmits the vital sign data of the patient to the cloud in priority according to the transmission strategy, and transmits the medical image data and voice recording data after the network condition is relieved.

[0038] In step 40, the vital sign data, the medical image data, and the voice recording data are analyzed in association to obtain a computing task and a task complexity, and the computing task is processed based on the task complexity and a cloud transmission strategy to determine a data processing result.

[0039] Optionally, after the cloud transmission strategy is determined and the original multi-modal data obtained, the first aid system, since the data of a single modality can not completely reflect the patient's condition, can more accurately grasp the condition through the association analysis between the multi-modal data, but the task complexity of different tasks is different, and the efficiency of the data processing manner is different when processing tasks of different complexities. Therefore, the first aid system can select a suitable data processing manner according to the actual situation to improve the overall processing capability, that is, the computing task and the corresponding task complexity are determined according to the association analysis between the vital sign data, the medical image data, and the voice recording data, such as disease diagnosis, condition prediction, etc. For details, see the description of steps 4011-4015.

[0040] Further, after the first aid system determines the computing task and the task complexity, the processing process of the computing task is allocated according to the cloud transmission strategy combined with the task complexity, the computing task is decomposed into a plurality of sub-tasks, and then the data is processed according to the load of the edge device itself and the network state combined with the sub-tasks, which can better utilize the computing capability of the edge device. For tasks that cannot be completed by the edge device, the data processed by the edge device and the remaining sub-tasks are processed in the cloud server, and finally the entire task processing process is completed to obtain the data processing result for the final data processing result to the first aid intelligent center and the mobile terminal, etc. For details, see the description of steps 4021-4024. The computing resources of the edge device and the cloud server are fully utilized to improve the computing efficiency and accuracy.

[0041] Further, in an embodiment, taking a patient with chest pain as an example, the edge device obtains vital sign data of the patient showing abnormal heart rate and fluctuating blood pressure, medical images showing possible stenosis of the coronary artery, and voice recordings of the patient's self-report of chest pain symptoms. Through association analysis, it is determined that the computing task is the diagnosis of coronary heart disease, and the task complexity is high. The computing task is transmitted to the cloud server, and the medical diagnosis model of the cloud is used for analysis to finally obtain the preliminary diagnosis result of coronary heart disease and treatment suggestions.

[0042] Step 50, the data processing result is transmitted to the emergency intelligent center and the mobile terminal in a hierarchical manner.

[0043] Optionally, the emergency system transmits the data processing result according to the importance of the data processing result and the use object. Specifically, for the emergency and critical diagnosis result and treatment suggestion, the highest priority is directly transmitted to the emergency intelligent center, so that the hospital emergency team can quickly develop a treatment plan; for the auxiliary information, such as the disease analysis process and related medical knowledge, the lower priority is transmitted to the mobile terminal (such as the mobile phone and tablet computer of the medical staff), so that the medical staff can check and refer at any time. During the transmission process, encryption technology is used to ensure the security and privacy of the data and prevent data leakage.

[0044] Further, the embodiment of the present application can enable different use objects to obtain the most needed information in time through hierarchical transmission, improve the use efficiency of information, help the emergency team to make decisions quickly, and at the same time ensure the safety of data and protect the privacy of patients.

[0045] Further, in an embodiment, taking an emergency patient with a sudden heart attack as an example, the data processing result shows that the patient needs to undergo a heart bypass surgery immediately, and this result is marked as high priority and is transmitted to the emergency intelligent center in the first time, so that the center can quickly allocate surgical resources and experts; at the same time, the brief disease and emergency measures are transmitted to the mobile terminal of the on-site medical staff to guide the on-site emergency work. The detailed disease analysis and follow-up rehabilitation suggestions of the patient are marked as low priority and are transmitted to the emergency intelligent center and the mobile terminal when the network is idle.

[0046] The embodiment of the present application ensures the efficiency of data transmission by adapting the protocol between the edge device and the 5G base station to obtain the target adaptation protocol corresponding to the edge device, avoids the compatibility problem between the device interfaces that are not unified; further, according to the transmission priority of the multi-modal data, the evaluation result of the 5G network and the task complexity at the edge device, the calculation task can be reasonably allocated, the local processing of simple tasks is reduced to reduce the delay, the complex tasks are processed in the cloud to ensure the accuracy of data transmission and processing, thereby improving the processing capacity of the edge device, avoiding the situation that the processing capacity of the edge device is insufficient when analyzing the data of complex conditions, and further ensuring that the emergency decision is not affected, ensuring the timeliness of emergency and improving the efficiency of emergency.

[0047] In an embodiment, steps 101-105 are described as follows:

[0048] Step 101, based on the analysis of the transmission information of the edge device to the 5G base station, a time domain signal is obtained.

[0049] Optionally, the emergency system uses a signal acquisition and receiving device installed on the edge device, such as a wideband spectrum analyzer. When the edge device receives the signal transmitted by the 5G base station, the analyzer monitors the signal transmitted by the 5G base station in real time, and after the received radio frequency signal is preprocessed by amplification, filtering, etc., it is converted into a processable electrical signal. Then, the amplitude, phase and other parameters of the electrical signal are analyzed with respect to time to obtain the time-domain signal. For example, the signal transmitted by the base station may contain carrier signals of different frequencies, and the receiving device converts these mixed signals into voltage or current signals that vary with time to obtain the time-domain signal.

[0050] Step 102, performing short-time Fourier transform on the time-domain signal to obtain a three-dimensional spectrum matrix, and extracting feature maps from the three-dimensional spectrum matrix to obtain initial spectrum fingerprint data.

[0051] Optionally, after determining the time-domain signal, the emergency system uses the short-time Fourier transform (STFT) time-frequency analysis method to transform the time-domain signal. Specifically, for the obtained time-domain signal, it is divided into several short time windows. In each time window, the signal is subjected to Fourier transform. The time-domain signal is decomposed into the superposition of sine and cosine waves of different frequency components to obtain the frequency component information of the signal in each time window. As the time window slides on the time-domain signal, a time-varying frequency information matrix, i.e. a three-dimensional spectrum matrix (time-frequency-power), is obtained.

[0052] Further, after obtaining the three-dimensional spectrum matrix, image recognition and feature extraction algorithms such as principal component analysis (PCA) and scale invariant feature transform (SIFT) are used to extract feature maps representing the unique characteristics of the signal from the matrix, and these feature maps are combined to form initial spectrum fingerprint data. For example, signals of different 5G communication protocols will have different distribution characteristics in the frequency spectrum, such as the power distribution of NSA and SA protocols in certain frequency bands. These differences can be captured through feature map extraction.

[0053] Step 103, matching the initial spectrum fingerprint data with the templates in the protocol fingerprint library to obtain similarity scores, and screening the similarity scores to obtain a preliminary adapted protocol.

[0054] Optionally, the emergency system is pre-provided with a protocol fingerprint library, which stores standard spectrum templates of various 5G communication protocols in different scenarios, and each template contains the corresponding spectrum feature parameter range of the protocol. After obtaining the initial spectrum fingerprint data, it is matched with the templates in the protocol fingerprint library. For example, Euclidean distance matching is performed to calculate the distance between the initial spectrum fingerprint data and each template in the feature parameter space. The closer the distance, the higher the similarity score. According to the set similarity threshold, the protocols with scores higher than the threshold are screened out to form a preliminary adaptive protocol set. Through matching with the protocol fingerprint library templates and similarity screening, the preliminary adaptive protocol that is consistent with the current signal characteristics can be quickly found from the numerous protocols, which narrows down the adaptation range and improves the efficiency of protocol adaptation, laying a foundation for subsequent accurate determination of the target adaptive protocol.

[0055] Step 104, based on the current network load and the edge device resource state, the preliminary adaptive protocol is screened again to obtain the candidate adaptive protocol.

[0056] Optionally, the emergency system monitors the current network load according to the edge device, such as monitoring the data flow, the number of connected devices and other indicators in the network to evaluate whether the network is in a busy state, and monitoring its own resource state, such as computing resources (such as CPU usage, memory occupancy), storage resources (remaining storage space), etc. According to each protocol in the preliminary adaptive protocol set, combined with the current network load and the edge device resource state, the secondary screening is performed. If the current network load is high, the protocol with low network resource demand is preferred; if the edge device computing capability is limited, the protocols with high computing resource requirements are excluded. Thus, it is ensured that the selected protocol can run stably and efficiently in the current actual environment, avoiding communication interruption or performance degradation due to mismatch between the protocol and the actual resources, improving the reliability and stability of communication.

[0057] Step 105, the performance of each protocol in the candidate adaptive protocol is evaluated to determine the target adaptive protocol.

[0058] Optionally, the emergency system establishes a test connection between the edge device and the 5G base station for each protocol in the candidate adaptation protocol in turn according to the obtained candidate adaptation protocol, monitors the response information of the base station including the time delay (i.e. the time required for the data packet to be sent from the edge device to the base station and returned), the packet loss rate (the proportion of unsuccessfully received data packets in the transmitted data packets), the data transmission rate and the like by sending a specific format of probe data packet, and judges whether the protocol meets the current communication demand according to the preset performance evaluation standard, such as requiring the time delay to be less than a certain threshold (such as 5 milliseconds), the packet loss rate to be lower than a certain proportion (such as 1%), the data transmission rate to reach a certain value (such as 100 Mbps) and the like. If yes, the next protocol is selected from the candidate adaptation protocol for verification until the optimal adaptation protocol, i.e. the target adaptation protocol, is found.

[0059] The embodiment of the present application fully and deeply mines the 5G base station signal characteristics through signal analysis, feature extraction, matching screening and performance evaluation, accurately identifies and adapts the protocol, reduces the possibility of misjudgment and mismatch, considers the network environment (such as network load) and the resource state of the edge device itself, can dynamically adjust the protocol adaptation process according to the actual situation, adapts to the complex and changeable 5G communication scene, and guarantees the stability of emergency communication. Finally, the protocol adaptation process is dynamically adjusted according to different network environments and device states, which adapts to the complex and changeable 5G communication scene, and has strong environmental adaptability.

[0060] In an embodiment, steps 201-204 are described as follows:

[0061] In step 201, the vital sign data, medical image data and voice recording data are parsed respectively, and feature extraction is performed based on the parsing results to obtain a data feature set.

[0062] Optionally, the emergency system first analyzes the heart rate, blood pressure, blood oxygen saturation and the like data collected by the sensor, such as analyzing the fluctuation frequency of the heart rate data and the change trend of the blood pressure. Through time series analysis and the like, real-time monitoring indicators and their change trends and the like are extracted. For CT, MRI and the like medical data images, image recognition technology is used to identify key lesion areas in the images, such as finding abnormal shadow areas in lung CT images; at the same time, image quality parameters such as image resolution, contrast and the like are extracted. These key lesion areas and image quality parameters constitute the features of the medical image data. For the obtained language recording data, speech recognition technology is used to convert the speech into text. Then natural language processing technology is used to extract diagnostic keywords such as "myocardial infarction", "fracture" and the like and emergency degree keywords such as "sudden" and "severe pain" and the like from the text. These keywords constitute the features of the voice recording data. Finally, the features extracted from the three types of data are integrated together to form a data feature set.

[0063] Step 202, judging each data feature in the data feature set based on the preset emergency degree evaluation library to obtain an emergency degree evaluation result.

[0064] Optionally, the emergency degree evaluation library pre-set in the emergency system stores emergency degree judgment standards corresponding to various data features. For vital sign data features, such as heart rate exceeding 160 beats per minute (for adults) or less than 40 beats per minute, the danger index threshold is used to judge as an emergency. For medical image data, if a large area of bleeding lesions is found in the key area of the brain, the lesion nature and location are evaluated as an emergency. For voice recording data, if emergency keywords such as "cardiac arrest" appear, semantic analysis is used to determine the emergency level. According to these emergency degree judgment standards, each data feature in the data feature set is judged one by one, and the data is divided into three levels of emergency, relatively emergency, and routine, to obtain an emergency degree evaluation result. The emergency degree evaluation has unified and clear standards, reduces the subjectivity and uncertainty of human judgment, and can quickly and accurately classify the emergency degree of the data to provide a reliable basis for subsequent priority determination.

[0065] Step 203, determining the timeliness weight of vital sign data, medical image data, and voice recording data based on the data type and clinical application scenario, and determining the timeliness score of each data based on the data collection time and timeliness weight.

[0066] Optionally, since the roles of different types of data in the emergency process vary with time, the emergency system determines the timeliness weight of vital sign data, medical image data, and voice recording data based on the data type and clinical application scenario. That is, for vital sign data, since it needs to be monitored in real time to monitor the patient's life state, it has a great influence on emergency decision-making, and the timeliness requirement is extremely high in the clinical application scenario, so it is given the highest timeliness weight; medical image data is crucial for determining the condition at the beginning of diagnosis, but its influence on subsequent treatment decisions decreases over time, so the timeliness weight is moderate; voice recording data has high reference value mainly in the recording of the condition, and the subsequent reference effect gradually decreases, so the timeliness weight is low, for example, the timeliness weight of vital sign data is set to 0.6, the timeliness weight of medical image data is set to 0.3, and the timeliness weight of voice recording data is set to 0.1. Then, the score is calculated according to the data collection time combined with the timeliness weight. Assuming that vital sign data is collected at the current time, denoted as t0, medical image data is collected at time t1 (t1 is 10 minutes away from t0), and voice recording data is collected at time t2 (t2 is 15 minutes away from t0). A time decay function is introduced, such as (where λ represents the decay constant, which can be set according to clinical experience). Then the timeliness score T of vital sign data is score1= 0.6 x 1 (current collection, no attenuation); medical image data timeliness score T score2 = 0.3 x e -λ×10 ; voice recording data timeliness score T score3 = 0.1 x e -λ×15 . And with the increase of time difference, the timeliness score gradually decreases.

[0067] Step 204, cross analysis of the emergency degree evaluation result and the timeliness score, to obtain the priority classification result.

[0068] Optionally, the emergency system first lists the data with the emergency degree of "emergency" as high priority, so as to realize direct priority processing and transmission of the data which has a great impact on the safety of patient's life. For the data with "more urgent" and "routine", combined with the timeliness score, the priority score is calculated by the formula , wherein n represents the number of data, E level represents the emergency degree level, T score represents the timeliness score, and then the final priority classification result is determined according to the calculated priority score. All data are sorted according to the priority score from high to low, and different priority categories are divided. Through cross analysis of the emergency degree and the timeliness, the importance and time value of the data can be comprehensively and comprehensively considered, so that the priority classification result is more scientific and reasonable.

[0069] The embodiment of the application comprehensively considers the key factors such as the emergency degree and the timeliness of the data, and can accurately determine the priority of the data, and provides reliable data support for emergency decision-making.

[0070] In an embodiment, steps 301-305 are described as follows:

[0071] Step 301, combine network bandwidth, delay, packet loss rate and node load resource index data with timestamp and node geographic location information of data collection, to obtain an original data set.

[0072] Optionally, in the 5G network, the edge device of the emergency system can monitor the network bandwidth, delay, packet loss rate and node load resource index data in real time, at the same time, the device also records the timestamp of these data collection for identifying the specific moment of data collection, and the built-in positioning module of the device also obtains the geographic location information of itself. Integrate these resource index data with the corresponding timestamp and geographic location information together to form an original data set containing resource index, time and space information.

[0073] Step 302, based on time intervals, the original data set is divided into time windows with time as the horizontal axis, and based on geographical regions, the edge devices are regionally divided with space as the vertical axis, and based on the time windows and different regions, a space-time two-dimensional window is constructed.

[0074] Optionally, the emergency system groups the data in the original data set according to fixed time intervals (such as 5 minutes) with time as the horizontal axis. Each group of data corresponds to a time window, so that each time window contains all the network resource indicator data of the edge devices collected in that time period. Then, according to geographical regions (such as different administrative districts of a city, different sub-regions of a park, etc.), the edge devices are classified and divided into different regions with space as the vertical axis. For example, a city can be divided into a central district, a suburban area, and other regions. Finally, the time windows and regions are cross-combined to form a space-time two-dimensional window. Each space-time two-dimensional window corresponds to a network resource indicator data set in a specific time and a specific region. By constructing a space-time two-dimensional window, the network resource data can be structured and organized in two dimensions of time and space. This facilitates subsequent targeted analysis of network resources in different times and regions, and mining of the change characteristics and rules of network resources in the space-time dimension.

[0075] Step 303, entropy value processing is performed on each resource data of the 5G network resources in each space-time window in the space-time two-dimensional window to obtain a resource entropy value, and each resource data of the 5G network resources in each space-time window in the space-time two-dimensional window is converted to obtain a resource indicator vector.

[0076] Optionally, when the emergency system performs entropy value processing on each resource data of the 5G network resources in each space-time window, taking network bandwidth as an example, for the bandwidth data x ij (i represents the time window number, j represents the region number), the probability of each data appearing in the window is calculated first (n represents the number of data in the window). Then the entropy value of the bandwidth in the window is calculated by the formula And the entropy values of delay, packet loss rate, node load and other resource indicators in the corresponding space-time window are calculated in the same way. The higher the entropy value, the greater the fluctuation of the corresponding resource indicator in the window, and the higher the uncertainty.

[0077] Further, after obtaining the corresponding resource entropy value, the network bandwidth, delay, packet loss rate, and node load resource data in each space-time window are standardized (such as normalized to the [0, 1] interval) and then arranged in a certain order to form a vector, i.e., a resource indicator vector. For example, the standardized bandwidth, delay, packet loss rate, and node load data are arranged in turn to obtain a resource indicator vector

[0078] In step 304, a space-time correlation matrix is constructed based on the similarity between resource index vectors in different space-time windows, and a space-time correlation degree is determined based on the space-time correlation matrix.

[0079] Optionally, the emergency system adopts a cosine similarity formula The similarity between resource index vectors in different space-time windows is calculated. Wherein A and B are resource index vectors of two space-time windows, a k and b k respectively represent the corresponding resource index values. All the similarity values between the calculated space-time windows are arranged in a certain order to construct a space-time correlation matrix. Each element in the matrix represents the similarity of the resource indexes of the corresponding two space-time windows. And by analyzing the space-time correlation matrix, the space-time correlation degree between different space-time windows is determined. The higher the similarity, the higher the space-time correlation degree.

[0080] In step 305, a resource evaluation result is determined based on the resource entropy value and the space-time correlation degree.

[0081] Optionally, the emergency system combines the resource entropy value and the space-time correlation degree to divide the network resource state into three levels: stable, general, and unstable. When the resource entropy value is low, it means that the resource index fluctuation is small, and the space-time correlation degree is also low, indicating that the resource states of different space-time windows are different and relatively stable. At this time, the network resource state is determined to be stable. When the resource entropy value is high, it means that the resource index fluctuation is large, and the space-time correlation degree is high, indicating that the resource states of different space-time windows are similar and unstable. It is determined to be an unstable state. In other cases, it is determined to be a general state. For example, the network resource state of a certain region in a certain period of time is stable, and the correlation degree with other space-time windows is also low.

[0082] The embodiment of the application comprehensively considers various characteristics of network resources, avoids the limitations of single dimension or single index analysis, and uses entropy calculation and cosine similarity scientific methods to quantify the fluctuation of resource indexes and the space-time correlation degree, so that the resource evaluation result is more accurate and reliable, and can truly reflect the actual state of network resources. Finally, by analyzing the space-time correlation degree, the change rule and trend of network resources in the space-time dimension are found, which helps to plan and adjust data transmission strategies in advance, improves the network resource utilization efficiency and the stability of the communication system.

[0083] In an embodiment, steps 4011-4015 are described as follows:

[0084] At step 4011, based on the vital sign data, a time series tree is constructed, and structural feature extraction is performed on the time series tree to obtain a first structural feature.

[0085] Optionally, since the vital sign data (such as heart rate, blood pressure, blood oxygen saturation, etc.) changes over time, the emergency system arranges the vital sign feature data collected at each time point in time sequence as a node, connects these nodes in time sequence, and constructs a time series tree. For example, heart rate data is collected every minute, and each collected heart rate value is taken as a node to form a time series tree reflecting the change of heart rate over time. Then, structural features are extracted from the time series tree, such as the number of branches, i.e. the number of child nodes extending from a node, which reflects the change branch situation of the vital sign data at the time point; the node depth, i.e. the number of edges from the root node to the node, which can reflect the depth information in time. Constructing the vital sign data into a time series tree and extracting the structural features can clearly present the change pattern and structural characteristics of the vital sign data in the time dimension.

[0086] At step 4012, based on the medical image data, a region topology graph is constructed, and graph structure feature extraction is performed on the region topology graph to obtain a second structural feature.

[0087] Optionally, since the medical image data (such as CT, MRI images) contains different anatomical regions and lesion information, the emergency system takes different regions in the image as nodes and the relationship (such as adjacency, inclusion, etc.) between regions as edges to construct a region topology graph. For example, in a lung CT image, different regions such as lung lobes, lung blood vessels, and lesions are taken as nodes, and connected into edges according to their spatial positional relationship in the image to form a region topology graph. Then, graph structure features are extracted from the region topology graph, such as node degree, i.e. the number of edges connected to a node, which reflects the association degree of the region in the entire image structure; the shortest path, i.e. the path with the fewest edges between two nodes, which can reflect the distance and close degree between different regions. By constructing the region topology graph and extracting the graph structure features, the spatial relationship and structural characteristics of different regions in the medical image can be intuitively displayed.

[0088] At step 4013, based on the voice recording data, a semantic hierarchy tree is constructed, and feature extraction is performed on the semantic hierarchy tree to obtain a third structural feature.

[0089] Optionally, the emergency system converts the voice recording data into text after voice recognition, and then performs semantic analysis on the text. The analyzed semantic units (such as words, phrases, sentences, etc.) are connected into a semantic hierarchical tree as nodes according to the hierarchical relationship of semantics (such as subordination, parallelism, etc.). For example, in a voice recording about the patient's condition, semantic units such as "patient", "symptom", "treatment", etc. can be used as nodes at different levels to build a semantic hierarchical tree. Then, the hierarchical depth is extracted from the semantic hierarchical tree, that is, the number of edges from the root node to a certain node, reflecting the nesting depth of the semantics; the node correlation degree can be obtained by calculating the closeness of the semantic correlation between nodes, such as calculating the semantic similarity by word vector.

[0090] Step 4014, similarity processing is performed between the first structural feature, the second structural feature and the third structural feature, to obtain a structural similarity score between any two structural features.

[0091] Optionally, the emergency system uses a graph edit distance algorithm to calculate the similarity between different types of data structures. For the time series tree of vital sign data and the semantic hierarchical tree of voice recording data, the structural similarity score of the two is obtained by calculating the number of add, delete and modify operations of the nodes. For example, when comparing the node structure reflecting the change of heart rate in the time series tree and the node structure describing the heartbeat abnormality semantics in the semantic hierarchical tree, if fewer node add, delete and modify operations are needed to make the two structures similar, it means that their similarity is high, and the score is also high. The specific formula is where G1 and G2 represent two data structures (such as time series tree and semantic hierarchical tree), m represents the number of operations; w i represents the operation weight (set according to the importance of different operations, such as higher weight for deleting important nodes), o i represents the cost of the i-th operation (such as insertion, deletion, modification) (which can be determined according to factors such as the difficulty of the operation). The same method is used to calculate the similarity scores between vital sign data and medical image data, medical image data and voice recording data structural features.

[0092] Step 4015, analyzing the structural similarity score to obtain the calculation task and the task complexity.

[0093] Optionally, the emergency system pre-sets a structural similarity threshold, when the similarity score of two data structures exceeds the threshold, it indicates that the two data have strong correlation in structure, and the joint analysis of the two data is divided into a computing task. For example, if the structural similarity score of the medical image region topology graph and the vital sign time series tree meets the standard, the comprehensive analysis of the patient's image and signs can be taken as a computing task. The complexity of the computing task is calculated by the formula where S represents the number of data types participating in the task, D i represents the average similarity distance of the kth data structure and other data structures. From the perspective of data structure difference, the more data types participating in the task and the greater the average similarity distance between the data structures, the higher the task complexity. Thus, the computing task and the task complexity are obtained.

[0094] The embodiment of the present application can deeply mine the internal structural relationship of data by constructing specific structures of different modal data and extracting features, and then calculating the structural similarity, which can more comprehensively and accurately find the correlation between data than traditional methods, and provide more valuable information for emergency decision-making. According to the structural similarity score, the computing task is divided, which makes the organization of the computing task more reasonable, avoids blind analysis, improves the efficiency of data processing and analysis, and helps to obtain meaningful results faster.

[0095] In an embodiment, steps 4021-4024 are described as follows:

[0096] In step 4021, based on the resource parameters of the edge device, the task load is determined, and based on the task complexity, the task load and the cloud transmission strategy, a hierarchical task strategy is constructed.

[0097] Optionally, the edge device of the emergency system first determines the task load that the edge device can bear according to its own resource parameters, such as computing power (measured by the number of instructions executed per second), storage capacity, network bandwidth, etc. According to these resource parameters, the task load that the edge device can bear is determined. For example, if the computing power of the edge device is limited and it can only process a certain number of computing instructions per second, the task load it can bear can be determined by combining the estimated computing amount of the task. Then, according to the task complexity, the higher the task complexity, the greater the computing amount and the higher the difficulty. Then, combined with the obtained cloud transmission strategy, a hierarchical task strategy is constructed. Specifically, steps 40211-40214 are described. According to the resource parameters of the edge device, the task load is determined, which can fully utilize the capabilities of the edge device and avoid resource waste or overload. The hierarchical task strategy is constructed, which combines the task complexity and the cloud transmission strategy, which can reasonably allocate tasks to edge devices and cloud servers, improve the overall processing efficiency, and ensure efficient processing of tasks under different network and resource conditions.

[0098] Further, taking the edge device on an ambulance as an example, its computing power is 10 million instructions per second, and its storage capacity is 10 GB. For a simple analysis task of vital sign data, the estimated computing amount is 5 million instructions per second, and the storage requirement is 1 GB. According to the resource parameters, it can be determined that the task load is within the range that the edge device can bear. In combination with the task complexity (the complexity of this task is low) and the sufficient current network bandwidth (the cloud transmission strategy tends to preferentially use cloud resources), it is divided into a task layer that can be processed at the edge device; and for a complex analysis task of medical images, the computing amount and storage requirement far exceed the capacity of the edge device, and the task complexity is high, so it is divided into a task layer that needs to be processed in the cloud.

[0099] Step 4022, based on the hierarchical task strategy, the computing task is decomposed to obtain a task decomposition result, and the task decomposition result is sent to the cloud server and the edge device.

[0100] Optionally, the first aid system decomposes the computing task according to the hierarchical task strategy obtained in the previous step. For the tasks divided into the edge device processing layer, they are further subdivided into multiple subtasks to ensure that each subtask can be efficiently executed within the resource range of the edge device. For example, the simple analysis task of vital sign data can be divided into a data cleaning subtask and a basic statistical analysis subtask. For the tasks divided into the cloud processing layer, they are also reasonably decomposed, such as the complex analysis task of medical images can be divided into an image segmentation subtask and a lesion recognition subtask. After decomposition, the task decomposition result (including the content, execution order, etc. of the subtasks) is sent to the cloud server and the edge device to make them clear about the tasks they need to process.

[0101] Step 4023, based on the edge device, the multi-modal data is preprocessed according to the task decomposition result to obtain a preprocessing result, and the preprocessing result is sent to the cloud server, so that the cloud server integrates and analyzes the preprocessing result based on the task decomposition result to obtain a task processing result, and returns the task processing result to the edge device.

[0102] Optionally, after receiving the task decomposition result, the edge device in the emergency system preprocesses the multi-modal data according to the sub-task assigned to itself. For example, for vital sign data, filtering processing can be performed to remove noise, and normalization processing can be performed to make the data within a suitable range; for medical image data, image enhancement operations can be performed to improve contrast, and cropping operations can be performed to remove irrelevant areas. After preprocessing, the preprocessing result is sent to the cloud server. The cloud server receives the preprocessing result sent by the edge device according to the task decomposition result, and integrates and analyzes the preprocessing result. For example, the medical image data preprocessed by the edge device is integrated with other related data (such as medical record information), and the powerful computing resources and deep learning models of the cloud are used to accurately identify the lesion and analyze the disease, to obtain the task processing result, and then return the result to the edge device.

[0103] The embodiment of the present application can reduce the amount of data and complexity by preprocessing through the edge device, reduce the processing pressure of the cloud server, and at the same time improve the processing speed by using the local processing capability of the edge device. The cloud server integrates and analyzes, which can utilize its powerful computing and storage resources to realize complex data analysis and model calculation, and ensure the accuracy and comprehensiveness of the analysis result. The cooperation of the two improves the efficiency and quality of the whole task processing.

[0104] Step 4024, combining the task processing result with the multi-modal data based on the edge device to determine the data processing result.

[0105] Optionally, after receiving the task processing result returned by the cloud server, the edge device in the emergency system combines it with the original multi-modal data. For example, the disease diagnosis result given by the cloud server is associated with the patient's vital sign data, medical image data, voice recording data, etc. to form a complete data processing result. The diagnosis result can be marked on the medical image, or the change trend of the vital sign data can be compared and analyzed with the diagnosis result, and finally a comprehensive and intuitive data processing result is presented, which is convenient for medical staff to view and understand.

[0106] The embodiment of the present application determines the task load and constructs a hierarchical task strategy according to the resource parameters of the edge device, reasonably allocates tasks to the edge device and the cloud server, fully utilizes the resource advantages of the two, avoids resource waste and overload, and improves the resource utilization efficiency; and through task decomposition, the complex task is refined, which is convenient for the edge device and the cloud server to efficiently execute respectively. The preprocessing of the edge device and the integration and analysis of the cloud server work together to speed up the task processing speed and improve the overall processing efficiency; finally, according to the cooperation of the edge device and the cloud server, the local processing advantage of the edge device is utilized, and the powerful computing and analysis capability of the cloud server is also utilized, to ensure the accuracy of the task processing result.

[0107] In an embodiment, the steps 42011-42014 are described as follows:

[0108] Step 42011, based on the importance, urgency and business needs of the computing task, determine the initial priority.

[0109] Optionally, the emergency system first analyzes the importance of the computing task to the emergency service for the computing task. For example, diagnostic tasks related to the safety of patients, such as diagnosis of acute myocardial infarction, are directly related to the safety of patients, and are crucial to the subsequent treatment and life protection of patients, so their importance is high; while some routine medical data statistical tasks, such as statistical tasks of the number of patients in a department in a week, their importance is relatively low. Then consider the urgency of the task, such as the emergency related computing task of a patient with sudden cardiac arrest, which needs to be handled immediately, and its urgency is high; while some regular analysis of the condition of patients with chronic diseases, its urgency is low. Finally, combined with business needs, such as computing tasks related to the current key research projects of the hospital, even if the urgency is not high, but due to business needs, it can also be given a certain priority. Combining these three factors, the initial priority of the computing task is divided into three levels: high, medium and low. For example, the diagnosis of acute myocardial infarction is set to high priority, the routine condition analysis task is set to medium priority, and the ordinary data statistical task is set to low priority.

[0110] Step 42012, based on the task complexity and task load, the initial priority is corrected to obtain the target priority.

[0111] Optionally, the emergency system combines the task complexity value, if the task complexity is high, it means that more computing resources and time are needed to process. For example, the complexity of complex medical image three-dimensional reconstruction and analysis task is high. When the task load of each node (such as edge device, cloud server) in the system is low, it means that there are enough resources to process high complexity tasks, at this time in order to ensure that important high complexity tasks can be processed first, its priority can be appropriately increased; on the contrary, if the task complexity is low, such as simple vital sign data mean calculation task, and the node load is high, the resource is tight, the priority can be reduced. By combining the task complexity and the task load, the initial priority is adjusted to obtain the target priority that is more consistent with the actual resource situation. For example, the high complexity task with the initial priority of medium can be corrected to high priority when the node load is low.

[0112] Step 42013, based on the cloud transmission strategy and the target priority of the computing task, the priority association relationship is constructed.

[0113] Optionally, the emergency system establishes an association relationship with the cloud transmission strategy according to the target priority of the computing task. For the task with high target priority, the delay priority transmission strategy is preferred to ensure that the task can be quickly transmitted to the processing node. For example, the vital sign data transmission task of an acute critical patient needs to be quickly transmitted to the cloud for analysis, and the delay priority transmission strategy is adopted. For the task with low target priority, the bandwidth priority transmission strategy can be adopted to fully utilize the network bandwidth resources, such as batch uploading of some historical medical data.

[0114] In step 42014, a hierarchical task queue is constructed based on the target priority of the computing task, and a hierarchical task strategy is constructed based on each queue in the hierarchical task queue, the task complexity, the task carrying capacity, and the priority association relationship.

[0115] Optionally, the emergency system divides the tasks into high-priority queues, medium-priority queues, and low-priority queues according to the target priority of the computing task, and constructs a hierarchical task queue. For example, tasks with high target priority are placed in the high-priority queue and wait for processing. Then, for each queue in the hierarchical task queue, a hierarchical task strategy is constructed in combination with the task complexity, the task carrying capacity, and the priority association relationship. For the tasks in the high-priority queue, since the priority is high, the node with sufficient resources and meeting the transmission strategy requirements is preferred when allocating the processing node and the transmission strategy. If the task complexity is high, the cloud server is preferred for processing, and the delay priority transmission strategy is adopted according to the priority association relationship. If the task complexity is low, the edge device or the cloud server can be reasonably allocated according to the node carrying capacity, while the associated transmission strategy is followed. The tasks in the medium-priority and low-priority queues are also allocated and processed in a similar manner by comprehensively considering various factors.

[0116] Further, in an embodiment, the high-priority queue has a diagnosis calculation task for an acute cerebral infarction patient, and the task complexity is high. According to the task carrying capacity, it is found that the current cloud server resource is sufficient, and the task is allocated to the cloud server for processing in combination with the priority association relationship (a delay priority transmission strategy is adopted), and the data is transmitted in a low-delay manner. The medium-priority queue has a routine examination report analysis task for a general patient, and according to the task complexity and carrying capacity, the task is allocated to the edge device for preliminary processing, and a suitable transmission strategy is adopted. The low-priority queue has a task such as a historical medical data backup task, which is processed when the system resource is idle.

[0117] The embodiment of the application determines the initial priority from the importance, urgency and business demand of the task, and then modifies the initial priority according to the task complexity and the task carrying capacity, comprehensively considers the task characteristics and the system resource status, and makes the priority determination more accurate and reasonable. By establishing the association relationship between the cloud transmission strategy and the task priority, and constructing the hierarchical task queue and the strategy, the processing node and the transmission strategy can be reasonably allocated according to the task priority and characteristics, the system resource can be optimally configured, and the resource utilization efficiency can be improved. Finally, the construction of the hierarchical task queue and the strategy enables the orderly management of tasks of different priorities, ensures the priority processing of high-priority tasks, and at the same time, takes into account the medium and low-priority tasks, improves the overall efficiency of task processing and the reliability of the system, and has good adaptability in complex and variable task environments.

[0118] Further, the 5G communication and edge device-based emergency method system provided by the application will be described below. The 5G communication and edge device-based emergency method system described below can be correspondingly referred to the 5G communication and edge device-based emergency method described above.

[0119] Optionally, referring to Figure 2 , Figure 2 is a structural schematic diagram of the 5G communication and edge device-based emergency method system provided by the application. The 5G communication and edge device-based emergency method system includes:

[0120] A protocol adaptation module is configured to detect the communication protocol type of the 5G base station based on the edge device, and to adapt the protocol of the edge device and the 5G base station based on the detection result to determine a target adaptation protocol.

[0121] A priority classification module is configured to acquire multi-modal data based on the edge device according to the target adaptation protocol, and to classify the multi-modal data according to priority to obtain a priority classification result. The multi-modal data includes vital sign data, medical image data and voice recording data.

[0122] A strategy determination module is configured to evaluate 5G network resources of the edge device, determine a resource evaluation result, and determine a cloud transmission strategy based on the resource evaluation result and the priority classification result.

[0123] A data processing module is configured to perform correlation analysis on the vital sign data, the medical image data and the voice recording data, obtain a computing task and a task complexity, and process the computing task based on the task complexity and the cloud transmission strategy to determine a data processing result.

[0124] A data delivery module is configured to deliver the data processing result to the emergency intelligent center and the mobile terminal in a hierarchical manner.

[0125] The edge device and the 5G base station are protocol adapted to obtain a target adaptation protocol of the edge device, ensuring efficient data transmission and avoiding compatibility problems between device interfaces that are not unified. Further, according to the transmission priority of the multi-modal data, the evaluation result of the 5G network and the task complexity at the edge device, the computing task can be reasonably allocated to realize local processing of simple tasks to reduce delay and cloud processing of complex tasks to ensure the accuracy of data transmission and processing, thereby improving the processing capability of the edge device, avoiding insufficient processing capability of the edge device during data analysis of complex conditions, ensuring that emergency decision-making is not affected, ensuring the timeliness of emergency and improving the efficiency of emergency.

[0126] Please refer to Figure 3 , Figure 3 An embodiment of an electronic device provided by the present application is shown in the figure. Figure 3 As shown in the figure, the present application provides an electronic device 300, which includes a memory 310, a processor 320, and a computer program 311 stored in the memory 310 and executable on the processor 320.

[0127] The communication protocol type of the edge device to the 5G base station is detected, and the edge device and the 5G base station are protocol adapted based on the detection result to determine a target adaptation protocol.

[0128] The multi-modal data is obtained by the edge device according to the target adaptation protocol, and the multi-modal data is classified by priority to obtain a priority classification result. The multi-modal data includes vital sign data, medical image data and voice recording data.

[0129] The 5G network resources of the edge device are evaluated to determine a resource evaluation result, and a cloud transmission strategy is determined based on the resource evaluation result and the priority classification result.

[0130] Correlation analysis is performed on the vital sign data, medical image data and voice recording data to obtain a computing task and a task complexity, and the computing task is processed based on the task complexity and a cloud transmission strategy to determine a data processing result;

[0131] The data processing result is transmitted to the emergency intelligent center and the mobile terminal in stages.

[0132] Please refer to Figure 4 , Figure 4 The embodiment of the computer readable storage medium provided by the present application is provided. As shown in Figure 4 , the present embodiment provides a computer readable storage medium 400, which stores a computer program 311, and the computer program 311 is executed by a processor to realize the following steps:

[0133] The communication protocol type of the 5G base station is detected based on the edge device, and the edge device and the 5G base station are protocol adapted based on the detection result to determine a target adaptation protocol;

[0134] Based on the target adaptation protocol, the edge device acquires multi-modal data, and the multi-modal data is prioritized to obtain a priority classification result. The multi-modal data includes vital sign data, medical image data and voice recording data;

[0135] The 5G network resources of the edge device are evaluated to determine a resource evaluation result, and based on the resource evaluation result and the priority classification result, a cloud transmission strategy is determined;

[0136] Correlation analysis is performed on the vital sign data, medical image data and voice recording data to obtain a computing task and a task complexity, and the computing task is processed based on the task complexity and a cloud transmission strategy to determine a data processing result;

[0137] The data processing result is transmitted to the emergency intelligent center and the mobile terminal in stages.

[0138] On the other hand, the present application also provides a computer program product, which comprises a computer program, the computer program can be stored on a non-transient computer readable storage medium, and the computer program is executed by a processor, and the computer can execute the emergency method based on the 5G communication and the edge device provided by the above-mentioned methods, and the method comprises:

[0139] The communication protocol type of the 5G base station is detected based on the edge device, and the edge device and the 5G base station are protocol adapted based on the detection result to determine a target adaptation protocol;

[0140] The edge device obtains multi-modal data according to a target adaptation protocol, and performs priority classification on the multi-modal data to obtain a priority classification result; the multi-modal data includes vital sign data, medical image data, and voice recording data;

[0141] The 5G network resources of the edge device are evaluated to determine a resource evaluation result, and based on the resource evaluation result and the priority classification result, a cloud transmission strategy is determined;

[0142] The vital sign data, the medical image data, and the voice recording data are associatedly analyzed to obtain a computing task and a task complexity, and based on the task complexity and the cloud transmission strategy, the computing task is processed to determine a data processing result;

[0143] The data processing result is hierarchically transmitted to a first-aid intelligent center and a mobile terminal.

[0144] The system embodiments described above are only illustrative, wherein the units illustrated as separate components can or can not be physically separated, and the components illustrated as units can or can not be physical units, i.e., can be located in one place, or can be distributed on multiple network units. Part or all of the modules can be selected to achieve the purpose of the embodiment according to actual needs. Those skilled in the art can understand and implement without creative labor.

[0145] Through the description of the above embodiments, those skilled in the art can clearly understand that the embodiments can be realized by means of software and necessary general hardware platforms, and of course, can also be realized by hardware. Based on such understanding, the above technical solutions can be embodied in the form of software products, which can be stored in a computer readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and include a number of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute the methods of each embodiment or some parts of the embodiment.

[0146] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement to some technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A first aid method based on 5G communication and edge device, characterized in that, The method comprises the following steps: Detecting the communication protocol type of the edge device to the 5G base station, and performing protocol adaptation between the edge device and the 5G base station based on the detection result to determine the target adaptation protocol; Based on the edge device, the multi-modal data is obtained according to the target adaptation protocol, and the multi-modal data is classified by priority to obtain a priority classification result; the multi-modal data includes vital sign data, medical image data and voice recording data; The 5G network resources of the edge device are evaluated to determine a resource evaluation result, and based on the resource evaluation result and the priority classification result, a cloud transmission strategy is determined; The vital sign data, medical image data and voice recording data are associated and analyzed to obtain a computing task and a task complexity, including: Based on the vital sign data, a time series tree is constructed, and structural features of the time series tree are extracted to obtain a first structural feature; Based on the medical image data, a regional topology graph is constructed, and graph structure features of the regional topology graph are extracted to obtain a second structural feature; Based on the voice recording data, a semantic hierarchy tree is constructed, and features of the semantic hierarchy tree are extracted to obtain a third structural feature; The first structural feature, the second structural feature and the third structural feature are processed for similarity between each other to obtain a structural similarity score between any two structural features; The structural similarity score is analyzed to obtain the computing task and the task complexity; And based on the task complexity and the cloud transmission strategy, the computing task is processed to determine a data processing result; including: Based on the resource parameters of the edge device, a task load is determined, and based on the task complexity, the task load and the cloud transmission strategy, a hierarchical task strategy is constructed; Based on the hierarchical task strategy, the computing task is decomposed to obtain a task decomposition result, and the task decomposition result is sent to a cloud server and the edge device; Based on the edge device, the multi-modal data is preprocessed according to the task decomposition result to obtain a preprocessing result, and the preprocessing result is sent to the cloud server, so that the cloud server integrates and analyzes the preprocessing result based on the task decomposition result to obtain a task processing result, and the task processing result is returned to the edge device; Based on the edge device, the task processing result and the multi-modal data are combined to determine the data processing result; The data processing result is transmitted to a first aid intelligent center and a mobile terminal in stages. 2.The 5G communication and edge device based first aid method of claim 1, wherein, Based on the task complexity, the task load and the cloud transmission strategy, a hierarchical task strategy is constructed, including: Based on the importance, urgency and business demand of the computing task, an initial priority is determined; The initial priority is corrected based on the task complexity and the task load to obtain a target priority; Based on the cloud transmission strategy and the target priority of the computing task, a priority association relationship is constructed; Based on the target priority of the computing task, a hierarchical task queue is constructed, and based on each queue in the hierarchical task queue, task complexity, task carrying capacity and the priority association relationship, the hierarchical task strategy is constructed. 3.The 5G communication and edge device based first aid method of claim 1, wherein, The edge device detects the communication protocol type of the 5G base station, and based on the detection result, the edge device and the 5G base station are protocol adapted to determine the target adaptation protocol, including: The edge device analyzes the transmission information of the 5G base station to obtain a time domain signal; Performing short-time Fourier transform on the time domain signal to obtain a three-dimensional frequency spectrum matrix, and extracting a feature spectrum from the three-dimensional frequency spectrum matrix to obtain initial frequency spectrum fingerprint data; Matching the initial frequency spectrum fingerprint data with templates in a protocol fingerprint library to obtain a similarity score, and screening the similarity score to obtain a preliminary adaptation protocol; Based on the current network load and the edge device resource state, the preliminary adaptation protocol is secondary screened to obtain a candidate adaptation protocol; The performance of each protocol in the candidate adaptation protocol is evaluated to determine the target adaptation protocol. 4.The 5G communication and edge device based first aid method of claim 1, wherein, The priority classification of the multi-modal data includes: The vital sign data, the medical image data and the voice recording data are respectively parsed, and feature extraction is performed based on the parsing results to obtain a data feature set; Based on a pre-set urgency evaluation library, each data feature in the data feature set is judged to obtain an urgency evaluation result; Based on data types and clinical application scenarios, the time effectiveness weight of the vital sign data, the medical image data and the voice recording data is determined, and based on data acquisition time and time effectiveness weight, the time effectiveness score of each data is determined; The urgency evaluation result and the time effectiveness score are cross-analyzed to obtain a priority classification result. 5.The 5G communication and edge device based first aid method of claim 1, wherein, The 5G network resources include network bandwidth, delay, packet loss rate and node load, and the evaluation of the 5G network resources of the edge device to determine the resource evaluation result includes: Combining network bandwidth, delay, packet loss rate and node load resource index data with time stamp and node geographic location information of data collection to obtain an original data set; Based on time intervals, the original data set is divided into time windows based on time as the horizontal axis, and based on geographic regions, the edge device is regionally divided into different regions based on space as the vertical axis, and based on time windows and different regions, a two-dimensional space-time window is constructed; Each resource data of the 5G network resources in each space-time window in the two-dimensional space-time window is entropy processed to obtain a resource entropy value; and each resource data of the 5G network resources in each space-time window in the two-dimensional space-time window is converted to obtain a resource index vector; Based on the similarity between the resource index vectors in different space-time windows, a space-time correlation matrix is constructed, and based on the space-time correlation matrix, a space-time correlation degree is determined; Based on the resource entropy value and the space-time correlation degree, the resource evaluation result is determined.

6. A first aid method system based on 5G communication and edge device, characterized in that, The application is applied to the first aid method based on 5G communication and edge device as claimed in any one of claims 1 to 5; the first aid method based on 5G communication and edge device system comprises: A protocol adaptation module is configured to detect the communication protocol type of the edge device to the 5G base station, and to adapt the protocol of the edge device and the 5G base station based on the detection result to determine the target adaptation protocol; A priority classification module is configured to obtain multi-modal data from the edge device according to the target adaptation protocol, and to classify the multi-modal data by priority to obtain a priority classification result; the multi-modal data includes vital sign data, medical image data and voice recording data; A strategy determination module is configured to evaluate the 5G network resources of the edge device, determine a resource evaluation result, and determine a cloud transmission strategy based on the resource evaluation result and the priority classification result; A data processing module is configured to perform correlation analysis on the vital sign data, the medical image data and the voice recording data to obtain a computing task and a task complexity, including: Based on the vital sign data, a time series tree is constructed, and structural feature extraction is performed on the time series tree to obtain a first structural feature; Based on the medical image data, a regional topology graph is constructed, and graph structure feature extraction is performed on the regional topology graph to obtain a second structural feature; Based on the voice recording data, a semantic hierarchy tree is constructed, and feature extraction is performed on the semantic hierarchy tree to obtain a third structural feature; The first structural feature, the second structural feature and the third structural feature are processed for similarity between each other to obtain a structural similarity score between any two structural features; The structural similarity score is analyzed to obtain the computing task and the task complexity; And based on the task complexity and the cloud transmission strategy, the computing task is processed to determine a data processing result; including: Based on the resource parameters of the edge device, the task load is determined, and based on the task complexity, the task load and the cloud transmission strategy, a hierarchical task strategy is constructed; Based on the hierarchical task strategy, the computing task is decomposed to obtain a task decomposition result, and the task decomposition result is sent to the cloud server and the edge device; Based on the task decomposition result, the edge device pre-processes the multi-modal data to obtain a pre-processing result, and sends the pre-processing result to the cloud server, so that the cloud server integrates and analyzes the pre-processing result based on the task decomposition result to obtain a task processing result, and returns the task processing result to the edge device; Based on the edge device, the task processing result and the multi-modal data are combined to determine the data processing result; A data delivery module is configured to transmit the data processing result to the first aid intelligent center and the mobile terminal in stages.

7. An electronic device, comprising: A memory is configured to store a computer software program; A processor for reading and executing the computer software program, wherein the processor, when executing the computer software program, implements the method for first aid based on 5G communication and edge device according to any one of claims 1 to 5.

8. A non-transitory computer readable storage medium having stored therein a computer software program, characterized in that, The computer software program, when executed by the processor, implements the method for first aid based on 5G communication and edge device according to any one of claims 1 to 5.

Images