Respiratory tract infection detection data storage method and system based on artificial intelligence

Abstract

The invention discloses a respiratory tract infection detection data storage method and system based on artificial intelligence, and belongs to the technical field of data storage. In a data storage process, network flow data corresponding to a data storage system is collected in real time, and a preset attack detection artificial intelligence model is adopted to identify the network flow data; according to the method, the network security state in the data storage process is determined, and when the network security state in the data storage process is secure, an independent storage block is divided for encrypted respiratory tract infection detection data corresponding to each disease to carry out classified storage on the respiratory tract infection detection data. Compared with existing local database direct storage or paper storage, the system is higher in safety, respiratory tract infection detection data can be effectively prevented from being leaked, data safety of patients is guaranteed, and meanwhile, data tracing is easier for doctors due to classified storage.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of data storage, and particularly relates to a respiratory tract infection detection data storage method and system based on artificial intelligence. BACKGROUND

[0002] Respiratory tract infection is one of the common diseases, and its rapid and accurate detection is of great significance for disease prevention and treatment. Respiratory tract infection detection data includes patient basic information, symptom description, laboratory test results, etc. Data storage needs to ensure safety, standardization and traceability, so as to facilitate doctors to check, compare and analyze at any time. These data are of great significance for disease prevention and treatment, treatment plan development and scientific research, and are an important basis for improving medical level and ensuring public health safety. However, the existing detection data storage method has problems of non-standard data management, low storage efficiency and poor safety, and an efficient, safe and standardized respiratory tract infection detection data storage method and system is urgently needed. SUMMARY

[0003] The application provides a respiratory tract infection detection data storage method and system based on artificial intelligence, which solves the problems of insecurity and non-standardization of the prior art.

[0004] In one aspect, the application provides a respiratory tract infection detection data storage method based on artificial intelligence, comprising: obtaining respiratory tract infection detection data and encrypting the respiratory tract infection detection data to obtain encrypted respiratory tract infection detection data; In the data storage process, real-time network traffic data corresponding to the data storage system is collected, and a preset attack detection artificial intelligence model is used to identify the network traffic data to determine the network security state in the data storage process; When the network security state in the data storage process is safe, a separate storage block is divided for each encrypted respiratory tract infection detection data corresponding to a disease for classified storage of the respiratory tract infection detection data, and data storage is completed.

[0005] Further, obtaining respiratory tract infection detection data and encrypting the respiratory tract infection detection data to obtain encrypted respiratory tract infection detection data comprises: obtaining respiratory tract infection detection data; wherein the respiratory tract infection detection data includes text data and respiratory tract image data; using an asymmetric encryption algorithm to encrypt the respiratory tract infection detection data to obtain encrypted respiratory tract infection detection data.

[0006] Further, in the data warehousing process, the network traffic data corresponding to the data storage system is collected in real time, and a preset attack detection artificial intelligence model is used to identify the network traffic data to determine the network security state in the data storage process, including: In the data warehousing process, the network traffic data corresponding to the data storage system is collected in real time, and the network traffic data is constructed into an input format of a preset attack detection artificial intelligence model to obtain to-be-identified data; The to-be-identified data is input into the preset attack detection artificial intelligence model, the actual output of the preset attack detection artificial intelligence model is obtained, and the network security state in the data storage process is obtained; wherein the network security state includes network security or network insecurity.

[0007] Further, before identifying the network traffic data using the preset attack detection artificial intelligence model, further comprising: The hyperparameters of the attack detection artificial intelligence model are initialized to obtain a plurality of different individuals; wherein each individual includes all hyperparameters of the attack detection artificial intelligence model; Obtain a network traffic attack data set; wherein the network traffic attack data set includes sample network traffic feature data and network security state labels corresponding to the sample network traffic feature data; For any one individual, the sample network traffic feature data and the network security state labels corresponding to the sample network traffic feature data are used as data support to obtain an error function value corresponding to the individual; According to the error function values corresponding to all individuals, an optimal individual is obtained; For any one individual, the optimal individual is used as a basis, and an adaptive surrounding exploration method is used to calibrate the exploration position of the individual to obtain the individual after the exploration position calibration; For any one individual after the exploration position calibration, an adaptive cooperative exploration method is used to explore the local region of the individual to obtain the individual after the local region exploration; For any one individual after the local region exploration, an adaptive mutation selection method is used to explore the global region of the individual to obtain the individual after the global region exploration; Determine whether the training end condition is met, if yes, then determine the optimal individual according to the individual after the global region exploration, and use the hyperparameters contained in the optimal individual as the final hyperparameters of the attack detection artificial intelligence model to obtain the preset attack detection artificial intelligence model, otherwise return to the step of obtaining the optimal individual.

[0008] Further, for any one individual, the optimal individual is used as a basis, and an adaptive surrounding exploration method is used to calibrate the exploration position of the individual to obtain the individual after the exploration position calibration, including: Based on the current number of training iterations, the adaptive location exploration control factor is obtained as follows: ;in, This represents the current number of training iterations, and T represents the maximum number of training iterations. This represents an exponential function with base e. The adjustment parameter is expressed as a constant; For any given individual, the adaptive control factor is obtained as follows: ;in, Represents a random number between (0, 1). Indicates the first t During the training process, the first i Individual, Represents an individual The modulus, D, represents the total dimension of the hyperparameters of an individual. i =1,2,…,N, where N represents the total number of individuals; Based on the optimal individual, the exploration location calibration value is determined as follows: Where cos represents the cosine function, sign represents the sign-finding function, and when the adaptive control factor... When greater than or equal to 0, It is 1 in all other cases. =-1; Represents the optimal individual; Based on the aforementioned exploration location calibration quantification, the exploration location of each individual is calibrated, resulting in the following individuals after exploration location calibration: ;in, Indicates the individual after the exploration location is marked. , Represents an individual The location of the exploration is quantified.

[0009] Furthermore, for any individual after its exploration location has been marked, an adaptive collaborative exploration method is used to explore a local region of that individual, resulting in an individual after the local region exploration, including: The adaptive density factor is obtained as follows: ;in, Indicates the first t Adaptive density factor during the training process. This represents the maximum value of the adaptive density factor, and is set to 2.7; This represents the minimum value of the adaptive density factor and is set to 0. represents the attenuation factor, and b represents the control parameter of the adaptive density factor; Based on the adaptive density factor, local region exploration is performed on the individual, resulting in the following individual after local region exploration: ;in, denotes a random number between (0, 1), denotes a learning factor as a constant, denotes a random number between (0, 1), denotes the m-th individual after the local area exploration in the t-th training process, t denotes the m-th individual after the local area exploration in the t-th training process, k denotes the m-th individual after the local area exploration in the t-th training process, denotes other individuals except the m-th individual, denotes the m-th individual after the local area exploration in the t-th training process, denotes the m-th individual after the local area exploration in the t-th training process.

[0010] Further, for any one of the individuals after the local area exploration, an adaptive mutation selection method is used to perform global area exploration on the individual to obtain an individual after global area exploration, including: For any one of the individuals after the local area exploration, the individual mutation value is obtained as: ; wherein, denotes the m-th individual after the local area exploration in the t-th training process, denotes the corresponding individual mutation value of the m-th individual, denotes the upper limit individual, denotes the lower limit individual, denotes a random mutation factor between (0, 1); When the error function value of the individual mutation value is less than the error function value of the individual , the individual mutation value is taken as the m-th individual after the global area exploration, otherwise the original individual is taken as the m-th individual after the global area exploration.

[0011] Further, it further includes: obtaining an image auxiliary processing task input by medical staff; wherein the image auxiliary processing task includes respiratory tract image data in respiratory tract infection detection data and image data labels corresponding to the respiratory tract image data; An artificial intelligence model is used to construct an image auxiliary processing model, and the respiratory tract image data and the image data labels corresponding to the respiratory tract image data are used to train the image auxiliary processing model to obtain a trained image auxiliary processing model.

[0012] Further, when the network security state in the data storage process is safe, a separate storage block is divided for each encrypted respiratory tract infection detection data corresponding to the disease to store the respiratory tract infection detection data in a classified manner, including: ​​When the network security state in the data storage process is safe, the trained image auxiliary processing model is used to assist in processing the respiratory tract image data in the respiratory tract infection detection data to realize the image auxiliary processing task specified by the medical staff, and determine the target image data label corresponding to the respiratory tract image data. After the encrypted respiratory tract infection detection data is associated with the corresponding target image data label, the target data to be stored is obtained. Each encrypted respiratory tract infection detection data corresponding to a disease is divided into a separate storage block, and the target data to be stored is stored in the corresponding storage block to complete data warehousing.

[0013] On the other hand, the present application provides a respiratory tract infection detection data storage system based on artificial intelligence, comprising: a data preprocessing module, a warehousing process safety detection module and a data warehousing storage module. The data preprocessing module is used to obtain respiratory tract infection detection data, and after the respiratory tract infection detection data is encrypted, encrypted respiratory tract infection detection data is obtained. The warehousing process safety detection module is used to collect network traffic data corresponding to the data storage system in real time during the data warehousing process, and a preset attack detection artificial intelligence model is used to identify the network traffic data to determine the network security state in the data storage process. The data warehousing storage module is used to divide a separate storage block for each encrypted respiratory tract infection detection data corresponding to a disease when the network security state in the data storage process is safe, and the respiratory tract infection detection data is classified and stored to complete data warehousing.

[0014] The present application provides a respiratory tract infection detection data storage method and system based on artificial intelligence, which collects network traffic data corresponding to the data storage system in real time during the data warehousing process, and uses a preset attack detection artificial intelligence model to identify the network traffic data to determine the network security state in the data storage process. When the network security state in the data storage process is safe, a separate storage block is divided for each encrypted respiratory tract infection detection data corresponding to a disease, and the respiratory tract infection detection data is classified and stored. Compared with the existing local database direct storage or paper storage, the security is stronger, which can effectively prevent the respiratory tract infection detection data from being leaked, and ensures the data security of the patient. At the same time, the classified storage makes it easier for doctors to trace the data. BRIEF DESCRIPTION OF DRAWINGS

[0015] The accompanying drawings incorporated in and forming a part of the specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the application.

[0016] Figure 1 A flowchart of a respiratory tract infection detection data storage method based on artificial intelligence provided by an embodiment of the present application.

[0017] Figure 2 A structural diagram of a respiratory tract infection detection data storage system based on artificial intelligence provided by an embodiment of the present application.

[0018] The specific embodiments of the present application have been shown in the above-described drawings, and will be described in more detail hereinafter. These drawings and the written description are not intended to restrict the scope of the inventive concept in any way, but to illustrate the inventive concept to those skilled in the art by reference to specific embodiments. DETAILED DESCRIPTION

[0019] The exemplary embodiments will be described in detail herein with reference to the attached drawings. In the following description, the same numbers are used to indicate the same or similar components. The embodiments described in the following exemplary embodiments are not intended to represent all embodiments consistent with the present application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of the present application as detailed in the appended claims.

[0020] The embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0021] As Figure 1 shown, the embodiment of the present application provides a respiratory tract infection detection data storage method based on artificial intelligence, comprising: S11, obtaining respiratory tract infection detection data, and obtaining encrypted respiratory tract infection detection data after encrypting the respiratory tract infection detection data; For example, the encrypted text data and the encrypted image data are encrypted separately, and the encrypted respiratory tract infection detection data can be obtained.

[0022] It is worth noting that the method of the present application can be applied to the internal system of the hospital, so as to make the process of data access or storage of doctors more secure, and can also be applied to the cloud. When the respiratory tract infection detection data transmitted by the terminal is obtained from the cloud, it also needs to be encrypted and transmitted to ensure the security of data transmission.

[0023] S12, in the data storage process, real-time collection of network flow data corresponding to the data storage system, and identification of the network flow data by using a preset attack detection artificial intelligence model to determine the network security state in the data storage process; In the existing hospital data storage system, the data storage security is often controlled by account password authentication, which is easy to cause data attacks, resulting in patient data leakage and threatening data security. Therefore, the preset attack detection artificial intelligence model is adopted to identify the network flow data in the embodiment of the application, and the network security state in the data storage process is determined, which can effectively improve the security in the data storage process.

[0024] S13, when the network security state in the data storage process is safe, a separate storage block is divided for each encrypted respiratory tract infection detection data corresponding to the disease to store the respiratory tract infection detection data, and the data storage is completed.

[0025] When the network security state in the data storage process is a non-safe state (that is, a certain attack is identified), the illegal attack device can be limited to access or blacklisted, so as to effectively improve the data security. Therefore, only when the network security state in the data storage process is safe, the data is stored, and the security of the entire database is ensured.

[0026] In the embodiment of the application, the respiratory tract infection detection data is obtained, and the respiratory tract infection detection data is encrypted to obtain the encrypted respiratory tract infection detection data, which comprises: The respiratory tract infection detection data is obtained, and the respiratory tract infection detection data is encrypted to obtain the encrypted respiratory tract infection detection data, which comprises: The respiratory tract infection detection data is encrypted by using an asymmetric encryption algorithm to obtain the encrypted respiratory tract infection detection data.

[0027] It is worth noting that other encryption methods can also be used to encrypt the text data and image data respectively to obtain the encrypted respiratory tract infection detection data.

[0028] In the embodiment of the application, in the data storage process, the network flow data corresponding to the data storage system is collected in real time, and the preset attack detection artificial intelligence model is used to identify the network flow data to determine the network security state in the data storage process, which comprises: In the data storage process, the network flow data corresponding to the data storage system is collected in real time, and the network flow data is constructed into the input format of the preset attack detection artificial intelligence model to obtain the to-be-identified data. The to-be-identified data is input into the preset attack detection artificial intelligence model, the actual output of the preset attack detection artificial intelligence model is obtained, and the network security state in the data storage process is obtained; wherein the network security state comprises network security or network insecurity.

[0029] In the embodiment of the present application, before the preset attack detection artificial intelligence model is used to identify the network flow data, the method further comprises the following steps: The hyperparameters of the attack detection artificial intelligence model are initialized to obtain a plurality of different individuals; wherein each individual includes all the hyperparameters of the attack detection artificial intelligence model; Optionally, the attack detection artificial intelligence model can be set as a convolutional neural network, and the main optimization hyperparameters are connection weights. After random initialization of these hyperparameters, the hyperparameters are encoded into a vector to obtain an individual. After repeating the operation multiple times, a plurality of different individuals can be obtained.

[0030] A network flow attack data set is obtained; wherein the network flow attack data set includes sample network flow feature data and network security state labels corresponding to the sample network flow feature data; The network flow attack data set can be set as a CICIDS2018 data set, a KDD99 data set, and a series of existing data sets. These data sets often contain sample network flow feature data and network security state labels corresponding to the sample network flow feature data. Learning from these data sets can achieve network security identification.

[0031] For any individual, the sample network flow feature data and the network security state labels corresponding to the sample network flow feature data are used as data support to obtain the error function value (such as cross-entropy loss function value) corresponding to the individual; According to the error function values corresponding to all individuals, the optimal individual is obtained; For any individual, the optimal individual is used as the basis, and the adaptive surrounding exploration method is used to calibrate the exploration position of the individual to obtain the individual after the exploration position calibration; For any individual after exploration position calibration, the adaptive cooperative exploration method is used to explore the local region of the individual to obtain the individual after local region exploration; For any individual after local region exploration, the adaptive mutation selection method is used to explore the global region of the individual to obtain the individual after global region exploration; It is judged whether the training end condition (such as the number of training times reaching the maximum training times) is met. If yes, the optimal individual is re-determined according to the individual after global region exploration, and the hyperparameters contained in the optimal individual are used as the final hyperparameters of the attack detection artificial intelligence model to obtain the preset attack detection artificial intelligence model. Otherwise, return to the step of obtaining the optimal individual.

[0032] Existing technologies suffer from problems such as getting trapped in local optima, poor training accuracy, and unsatisfactory training results during hyperparameter training. Therefore, this invention provides a new algorithm to improve the accuracy of network security identification.

[0033] In this embodiment of the invention, for any individual, based on the optimal individual, an adaptive orbital exploration method is used to determine the individual's exploration position, resulting in an individual with the exploration position determined, including: Based on the current number of training iterations, the adaptive location exploration control factor is obtained as follows: ;in, This represents the current number of training iterations, and T represents the maximum number of training iterations. This represents an exponential function with base e. The adjustment parameter is expressed as a constant; For any given individual, the adaptive control factor is obtained as follows: ;in, Represents a random number between (0, 1). Indicates the first t During the training process, the first i Individual, Represents an individual The modulus, D, represents the total dimension of the hyperparameters of an individual. i =1,2,…,N, where N represents the total number of individuals; Based on the optimal individual, the exploration location calibration value is determined as follows: Where cos represents the cosine function, sign represents the sign-finding function, and when the adaptive control factor... When greater than or equal to 0, It is 1 in all other cases. =-1; Represents the optimal individual; Based on the aforementioned exploration location calibration quantification, the exploration location of each individual is calibrated, resulting in the following individuals after exploration location calibration: ;in, Indicates the individual after the exploration location is marked. , Represents an individual The location of the exploration is quantified.

[0034] The adaptive surrounding exploration method provided by the application can make each individual adaptively select a respective exploration area around the optimal position based on the individual position, and meanwhile realize adaptive exploration, so that all individuals are shrunk in a net shape, and the training speed and the exploration possibility of the global optimal solution are effectively improved.

[0035] In the embodiment of the application, for any individual after the exploration position is calibrated, the adaptive cooperative exploration method is used to perform local area exploration on the individual, to obtain an individual after the local area exploration, comprising: The adaptive density factor is obtained as follows: ; wherein, denotes the adaptive density factor in the i th training process, t denotes the maximum value of the adaptive density factor, and is set to 2.7; denotes the minimum value of the adaptive density factor, and is set to 0; denotes a decay factor, and b denotes a control parameter of the adaptive density factor; According to the adaptive density factor, the local area exploration is performed on the individual, to obtain an individual after the local area exploration as follows: ; wherein, denotes a random number between 0 and 1, denotes a constant learning factor, denotes a random number between 0 and 1, denotes the individual after the i th exploration position calibration in the i th training process, denotes other individuals except the individual x i, t denotes the individual after the local area exploration x i. k The adaptive cooperative exploration method provided by the application can make the individual explore the area between two individuals with a search step from large to small, which is more helpful to improve the possibility of exploring the global optimal solution after the net-shaped shrinking exploration process. And the search efficiency is maintained relatively large in the early and middle stages of the algorithm, and gradually changes to fine search in the later stage of the algorithm, so as to improve the search precision. In the embodiment of the application, for any individual after the local area exploration, the adaptive mutation selection method is used to perform global area exploration on the individual, to obtain an individual after the global area exploration, comprising:

[0036] The adaptive cooperative exploration method provided by the application can make the individual explore the area between two individuals with a search step from large to small, which is more helpful to improve the possibility of exploring the global optimal solution after the net-shaped shrinking exploration process. And the search efficiency is maintained relatively large in the early and middle stages of the algorithm, and gradually changes to fine search in the later stage of the algorithm, so as to improve the search precision.

[0037] In the embodiment of the application, for any individual after the local area exploration, the adaptive mutation selection method is used to perform global area exploration on the individual, to obtain an individual after the global area exploration, comprising: ​​​For any one local area exploration after the individual, the individual mutation value is obtained: ; Wherein, Indicates the mth local area exploration after the individual in the tth training process, Indicates the individual Corresponding individual mutation value, Indicates the upper limit individual, Indicates the lower limit individual, Indicates a random mutation factor between (0, 1) ; When the error function value of the individual mutation value Is less than the error function value of the individual , then the individual mutation value As the mth global area exploration after the individual, otherwise the original individual As the mth global area exploration after the individual.

[0038] The adaptive mutation selection method provided by the application further improves the ability to jump out of the local optimal solution, and the adaptive selection of offspring can ensure the exploration efficiency of the algorithm.

[0039] Through the cooperation of the above several ways, the global exploration ability is comprehensively improved, and the training speed and training accuracy of the algorithm are guaranteed, so that the data relationship learning effect is better, and finally the network security recognition accuracy is improved, and the safety of the data storage process is guaranteed.

[0040] In the embodiment of the application, further comprising: Obtaining an image auxiliary processing task input by medical staff; wherein the image auxiliary processing task includes respiratory tract image data in respiratory tract infection detection data and image data labels corresponding to the respiratory tract image data; An image auxiliary processing model is constructed using an artificial intelligence model, and the image auxiliary processing model is trained using the respiratory tract image data and the image data labels corresponding to the respiratory tract image data, to obtain the image auxiliary processing model after training.

[0041] For example, the image auxiliary processing model can be set as a convolutional neural network, so as to complete the auxiliary classification task formulated by the doctor. The image auxiliary processing model can also be set as a U-net neural network, so as to complete the auxiliary segmentation task formulated by the doctor, thereby improving the data utilization efficiency. However, it should be noted that the above two image auxiliary processing models are only examples of the embodiment of the application, and can also be set as other artificial intelligence models.

[0042] In the embodiment of the present application, when the network security state in the data storage process is safe, a separate storage block is divided for each encrypted respiratory tract infection detection data corresponding to the disease, and the respiratory tract infection detection data is classified and stored, including: When the network security state in the data storage process is safe, the trained image auxiliary processing model is used to assist in processing the respiratory tract image data in the respiratory tract infection detection data, so as to realize the image auxiliary processing task specified by the medical staff, and determine the target image data label corresponding to the respiratory tract image data. After the encrypted respiratory tract infection detection data and the corresponding target image data label are associated, the target data to be stored is obtained. Each encrypted respiratory tract infection detection data corresponding to the disease is divided into a separate storage block, and the target data to be stored is stored in the corresponding storage block, and the data storage is completed.

[0043] Optionally, the storage block can use the patient ID as the unique identifier, so as to realize the classified storage of the patient data, and the target image data label is also stored, which is more convenient for doctors to trace the stored data and improve the data searching and using efficiency.

[0044] The present application provides a kind of respiratory tract infection detection data storage method based on artificial intelligence, by real-time acquisition about data storage system corresponding network flow data in data storage process, and using the preset attack detection artificial intelligence model is identified to network flow data, determine the network security state in the data storage process, and when the network security state in the data storage process is safe, a separate storage block is divided for each encrypted respiratory tract infection detection data corresponding to the disease, and the respiratory tract infection detection data is classified and stored, compared with the direct storage of existing local database or paper storage, stronger security, can effectively prevent respiratory tract infection detection data from being leaked, guarantee the data security of patient, while classified storage makes doctor more easily data traceback.

[0045] As shown in Figure 2 The present application provides a kind of respiratory tract infection detection data storage system based on artificial intelligence, including: data preprocessing module 21, storage process safety detection module 22 and data storage module 23; The data preprocessing module 21 is used to obtain respiratory tract infection detection data, and after the respiratory tract infection detection data is encrypted, the encrypted respiratory tract infection detection data is obtained. The warehousing process security detection module 22 is used for collecting network flow data corresponding to a data storage system in real time during a data warehousing process, and identifying the network flow data by using a preset attack detection artificial intelligence model to determine a network security state in the data storage process; The data warehousing storage module 23 is used for, when the network security state in the data storage process is safe, dividing a single storage block for each encrypted respiratory tract infection detection data corresponding to a disease to store the respiratory tract infection detection data, and completing data warehousing.

[0046] The respiratory tract infection detection data storage system can execute the above method scheme, and has the same principle and effect, which will not be described here.

[0047] Other embodiments of the application will be apparent to those skilled in the art from consideration of the specification and practice of the application disclosed herein. It is intended that the specification and examples be considered as exemplary only, with the true scope of the application being indicated by the following claims.

Claims

1. A data storage method for respiratory infection detection based on artificial intelligence, characterized in that, The method comprises the following steps: Obtain respiratory tract infection detection data, and encrypt the respiratory tract infection detection data to obtain encrypted respiratory tract infection detection data; During the data storage process, network traffic data corresponding to the data storage system is collected in real time, and a preset attack detection artificial intelligence model is used to identify the network traffic data to determine the network security state during the data storage process; When the network security state during the data storage process is safe, a separate storage block is allocated for each encrypted respiratory tract infection detection data corresponding to a disease to store the respiratory tract infection detection data, and the data storage is completed. 2.The artificial intelligence-based respiratory infection detection data storage method of claim 1, wherein Obtain respiratory tract infection detection data, and encrypt the respiratory tract infection detection data to obtain encrypted respiratory tract infection detection data, comprising: Obtain respiratory tract infection detection data; wherein the respiratory tract infection detection data includes text data and respiratory tract image data; Encrypt the respiratory tract infection detection data using an asymmetric encryption algorithm to obtain encrypted respiratory tract infection detection data. 3.The artificial intelligence-based respiratory infection detection data storage method of claim 1, wherein During the data storage process, network traffic data corresponding to the data storage system is collected in real time, and a preset attack detection artificial intelligence model is used to identify the network traffic data to determine the network security state during the data storage process, comprising: During the data storage process, network traffic data corresponding to the data storage system is collected in real time, and the network traffic data is constructed into an input format of the preset attack detection artificial intelligence model to obtain identification data; The identification data is input into the preset attack detection artificial intelligence model to obtain the actual output of the preset attack detection artificial intelligence model, and the network security state during the data storage process is obtained; wherein the network security state includes network security or network insecurity. 4.The artificial intelligence-based respiratory infection detection data storage method of claim 3, wherein, Before identifying the network traffic data using the preset attack detection artificial intelligence model, the method further comprises the following steps: Initialize the hyperparameters of the attack detection artificial intelligence model to obtain a plurality of different individuals; wherein each individual includes all hyperparameters of the attack detection artificial intelligence model; Obtain a network traffic attack data set; wherein the network traffic attack data set includes sample network traffic feature data and network security state labels corresponding to the sample network traffic feature data; For any one individual, use the sample network traffic feature data and the network security state labels corresponding to the sample network traffic feature data as data support to obtain an error function value corresponding to the individual; According to the error function values corresponding to all individuals, obtain an optimal individual; For any one individual, use the optimal individual as a basis to perform exploration position calibration on the individual using an adaptive surrounding exploration method to obtain an individual after exploration position calibration; For any one individual after exploration position calibration, use an adaptive cooperative exploration method to perform local area exploration on the individual to obtain an individual after local area exploration; For any one individual after local area exploration, use an adaptive mutation selection method to perform global area exploration on the individual to obtain an individual after global area exploration; If the training end condition is met, the optimal individual is determined again based on the individual after global region exploration, and the hyperparameters included in the optimal individual are used as the final hyperparameters of the attack detection artificial intelligence model to obtain a preset attack detection artificial intelligence model. 5.The artificial intelligence-based respiratory infection detection data storage method of claim 4, wherein For any individual, the optimal individual is used as the basis to explore the position of the individual using an adaptive surrounding exploration method to obtain an individual after position calibration, including: According to the current training number, an adaptive position exploration control factor is obtained as: ; wherein, represents the current training number, T represents the maximum training number, represents an exponential function with a natural constant e as the base, represents a regulation parameter that is a constant; For any one individual, the adaptive control factor is obtained as: ; wherein, represents a random number between (0, 1), represents the first t training process, i the first individual, the modulus of the individual, D represents the total dimension of the hyperparameters of the individual, i = 1, 2, …, N, and N represents the total number of individuals; According to the optimal individual, the exploration position calibration quantity is determined as: ; wherein, cos represents a cosine function, sign represents a sign function, when the adaptive control factor is greater than or equal to 0, is 1, and otherwise is -1; represents the optimal individual; According to the exploration position calibration quantity, the individual is calibrated in an exploration position, and the individual after exploration position calibration is obtained as: ; wherein, represents the individual after exploration position calibration , represents the exploration position calibration quantity of the individual . 6.The artificial intelligence-based respiratory infection detection data storage method of claim 5, wherein, For any individual after position calibration, the individual is explored in the local region using an adaptive cooperative exploration method to obtain an individual after local region exploration, including: The adaptive density factor is obtained as: ; wherein, denotes the adaptive density factor in the nth training process, t denotes the maximum value of the adaptive density factor and is set to 2.7; denotes the minimum value of the adaptive density factor and is set to 0; denotes a decay factor, and b denotes a control parameter of the adaptive density factor.​ Based on the adaptive density factor, local region exploration is performed on the individual, resulting in the following individual after local region exploration: ;in, Represents a random number between (0, 1). The learning factor is expressed as a constant. Represents a random number between (0, 1). Indicates the first t During the training process, the first k An individual after its location has been determined. Including individuals Other individuals, Indicates an individual after exploring a local area. . 7.The artificial intelligence-based respiratory infection detection data storage method of claim 6, wherein, For any individual after local region exploration, the individual is explored in the global region using an adaptive mutation selection method to obtain an individual after global region exploration, including: For any individual after local region exploration, the individual mutation value is obtained as: ; wherein, represents the individual after the mth local region exploration in the tth training process, represents the individual corresponding individual mutation value, represents the upper limit individual, represents the lower limit individual, represents a random mutation factor between (0, 1); When the error function value of the individual is less than the error function value of the individual , the individual mutation value is taken as the individual after the mth global region exploration, otherwise the original individual is taken as the individual after the mth global region exploration. 8.The artificial intelligence-based respiratory infection detection data storage method of claim 1, wherein, Also including: An image-assisted processing task input by medical staff is obtained, wherein the image-assisted processing task includes respiratory tract image data in respiratory tract infection detection data and image data labels corresponding to the respiratory tract image data; An image-assisted processing model is constructed using an artificial intelligence model, and the image-assisted processing model is trained using the respiratory tract image data and the image data labels corresponding to the respiratory tract image data to obtain a trained image-assisted processing model. 9.The artificial intelligence-based respiratory infection detection data storage method of claim 8, wherein, When the network security state in the data storage process is safe, a separate storage block is divided for each encrypted respiratory tract infection detection data corresponding to a disease to store the respiratory tract infection detection data, including: When the network security state in the data storage process is safe, the trained image-assisted processing model is used to assist in processing the respiratory tract image data in the respiratory tract infection detection data to achieve the image-assisted processing task specified by the medical staff, and determine the target image data label corresponding to the respiratory tract image data; After associating the encrypted respiratory tract infection detection data with the corresponding target image data label, the target data to be stored is obtained; Each encrypted respiratory tract infection detection data corresponding to a disease is divided into a separate storage block, and the target data to be stored is stored in the corresponding storage block to complete data warehousing.

10. An artificial intelligence based respiratory infection detection data storage system, characterized by, Including: A data preprocessing module, a storage process security detection module, and a data warehousing storage module; The data preprocessing module is configured to obtain respiratory tract infection detection data, encrypt the respiratory tract infection detection data, and obtain encrypted respiratory tract infection detection data; The storage process security detection module is configured to, during data warehousing, collect network traffic data corresponding to a data storage system in real time, and identify the network traffic data using a preset attack detection artificial intelligence model to determine the network security state in the data storage process; The data warehousing storage module is configured to, when the network security state in the data storage process is safe, divide a separate storage block for each encrypted respiratory tract infection detection data corresponding to a disease to store the respiratory tract infection detection data, and complete data warehousing.

Images