A cloud-based collaborative medical data management system and its use method

Abstract

The present invention relates to the field of management technology, and specifically discloses a cloud-based collaborative medical data management system and a method for using the same, including an image data acquisition module, a cloud storage module, a security protection module, and a collaborative management module. The present invention collects prostate ultrasound image data from personal devices of medical institutions and doctors through the image data acquisition module and uploads it to the cloud storage module. The user authority management module provides user identity authentication, authority management, and behavior recording functions. The security protection module adopts end-to-end encryption, and then uses a gradient descent algorithm through the collaborative management module to obtain the optimal solution of the objective function of the real-time collaborative operation queue model, thereby realizing efficient cloud-based collaborative processing in the prostate ultrasound image data management system.

Description

Technical Field

[0001] The present invention relates to the field of management technology, and more specifically, to a cloud-based collaborative medical data management system and a method for using the same. Background Art

[0002] Traditional prostate ultrasound image data management systems usually use local storage and processing methods, which have limited storage space and difficulties in data sharing. There is a lack of effective technical means for cross-departmental and cross-institutional data collaboration, which easily leads to information silos and affects the efficient use of prostate ultrasound image data. Although existing technologies have begun to try to introduce cloud computing technology to improve the efficiency of prostate ultrasound image data management, there is a problem of low efficiency in the cloud-based collaborative processing process in the prostate ultrasound image data management system.

[0003] In order to solve the above problems, a technical solution is now provided. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a cloud-based collaborative medical data management system and a method for using the same. The image data acquisition module collects prostate ultrasound image data from medical institutions and doctors' personal devices and uploads it to a cloud storage module. The user authority management module provides user authentication, authority management and behavior recording functions. The security protection module adopts end-to-end encryption to ensure that the data is always encrypted throughout the entire process from the sending end to the receiving end. Collaborative management uses a gradient descent algorithm to obtain the optimal solution of the objective function of the real-time collaborative operation queue model to solve the problems raised in the above background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A cloud-based collaborative medical data management system includes an image data acquisition module, a cloud storage module, a data processing module, a user rights management module, a security protection module, and a collaborative management module. The image data acquisition module acquires prostate ultrasound image data and uploads it to the cloud storage module. The data processing module includes a data analysis unit and an image processing function unit, supporting parallel data processing. The user rights management module provides user authentication, rights management, and behavior recording functions. The security protection module uses end-to-end encryption to ensure that data is always encrypted throughout the entire process from the sender to the receiver. The collaborative management module uses a gradient descent algorithm to obtain the optimal solution to the objective function of a real-time collaborative operation queue model, thereby achieving efficient cloud-based collaborative processing in the cloud-based collaborative medical data management system. The formula for the objective function of the real-time collaborative operation queue model is:

[0007] ,

[0008] Where: The objective function of the real-time collaborative operation queue model is is the weight coefficient of average response time, is the throughput weight coefficient, is the weight coefficient of memory resource utilization, is the total number of data sharing and real-time collaborative operations, Index variables for data sharing and real-time collaborative operations, is the average response time, is the weight associated with the dependency between data sharing and real-time collaborative operations, For the data sharing and real-time collaborative operation response time, is the throughput, The total time required to complete all data sharing and real-time collaborative operations, For the The contribution factors of data sharing and real-time collaborative operations to model throughput, is the memory resource utilization, is the total number of memory resource types, is the memory resource type, Memory resource type The total amount, For the Data sharing and real-time collaborative operation on memory resources The specific demand, For collaborative operation queue;

[0009] Using the gradient descent algorithm, the objective function The partial derivative with respect to the average response time is , the objective function The partial derivative with respect to throughput is , the objective function The partial derivative of memory resource utilization is , the formula for updating variables according to the gradient descent method is:

[0010]

[0011]

[0012] ,

[0013] Where: is the learning rate, set to 0.01, No. Response time for specific data sharing and real-time collaborative operations, After the update data sharing and real-time collaborative operation response time, The total time required to complete all data sharing and real-time collaborative operations, The total time required to complete all data sharing and real-time collaboration operations after the update. For the Data sharing and real-time collaborative operation on memory resources The specific demand, After the update Data sharing and real-time collaborative operation on memory resources The specific demand, repeat the above steps, constantly update the variables, and detect the objective function The rate of change in each iteration is dynamically determined to stop the iteration when the objective function The rate of change is less than for 10 consecutive iterations , stop iteration, objective function The optimal solution was reached.

[0014] As a further solution of the present invention, the image data acquisition module is connected to the cloud storage module, the data processing module is connected to the cloud storage module, the user authority management module is connected to the data processing module, the security protection module is connected to the user authority management module, and the collaborative management module is connected to the security protection module.

[0015] As a further solution of the present invention, the image data acquisition module collects prostate ultrasound image data from different medical institutions and doctors' personal devices and uploads it to the cloud, including:

[0016] The image data acquisition module connects to the medical information systems of various institutions and adopts the HL7 international medical information exchange standard to stably obtain prostate ultrasound image data from the servers of various medical institutions. For doctors' personal devices, the image data acquisition module provides a dedicated application that supports Windows, iOS, and Android operating systems. Doctors can install the dedicated application to upload prostate ultrasound image data stored on their personal devices to the image data acquisition module.

[0017] The collected prostate ultrasound image data has three formats: DICOM, JPEG and PNG. The image data acquisition module converts the three different formats of prostate ultrasound image data into a unified format, and uses the OpenCV open source image processing library to convert the prostate ultrasound image from JPEG format and PNG format to DICOM format. The prostate ultrasound image data converted into the unified DICOM format is set to be automatically uploaded to the cloud at 2 am every day. During the upload process, the upload progress is displayed in real time, including the amount of uploaded data and the remaining time information. The user can understand the progress of the upload through the progress bar. After the upload is completed, the image data acquisition module uses a hash algorithm to perform integrity verification on the uploaded data to ensure that the data is not lost or damaged during transmission.

[0018] As a further solution of the present invention, the cloud storage module provides the ability to store data and is composed of a distributed database unit and a cloud storage technology unit, including:

[0019] The distributed database unit achieves high data availability and fault tolerance by distributing prostate imaging data across multiple servers and nodes. The distributed database unit also includes data sharding, data replication, data consistency, availability, and partition tolerance. Data sharding divides data into multiple parts based on patient ID, with each part stored on a different node. To improve data fault tolerance and availability, data is typically replicated to multiple nodes. Distributed databases require a balance between consistency, availability, and partition tolerance.

[0020] The cloud storage technology unit provides elastic, scalable, and highly available storage solutions. Cloud storage is combined with object storage services to manage large-scale unstructured data, such as prostate ultrasound images and patient information. The cloud storage technology unit stores data in the form of objects in a logically unified storage pool and uses redundancy technology to ensure high data availability.

[0021] As a further solution of the present invention, the data processing module includes a data analysis unit and an image processing functional unit, supporting parallel processing of large-scale data, including:

[0022] The data analysis unit performs multi-dimensional quantitative analysis on the prostate ultrasound images, reduces the noise of the prostate ultrasound images through median filtering, performs equalization on the grayscale histogram of the prostate ultrasound images to enhance the image contrast, uses the Canny edge detection algorithm to extract the prostate boundary information in the prostate ultrasound images, and uses the deep learning-based segmentation algorithm U-Net to identify and segment nodules in the prostate ultrasound images and annotate them. By performing three-dimensional reconstruction on multiple frames of prostate ultrasound images, the three-dimensional structure of the prostate is reconstructed to provide intuitive visualization information and establish a multi-frame ultrasound three-dimensional pixel reconstruction model. The formula of the multi-frame ultrasound three-dimensional pixel reconstruction model is:

[0023] ,

[0024] Where: is the horizontal position of the prostate ultrasound image, is the vertical position of the prostate ultrasound image, is the depth coordinate in three-dimensional space, representing the The position of the prostate ultrasound image frame in three-dimensional space, In three-dimensional space, the coordinates Determine a three-dimensional pixel value, After image registration processing The pixel value of the frame prostate ultrasound image on the image plane, where , is a two-dimensional affine transformation matrix, For the Pixel values ​​in the frame 2D prostate ultrasound image;

[0025] The image processing functional unit uses a multi-core processor to decompose image processing tasks into multiple threads for parallel execution, thereby improving processing efficiency. Through stream processing technology, it processes prostate ultrasound image data in real time and provides processing results in a timely manner. It uses GPUs for large-scale matrix operations to accelerate the execution of image processing algorithms.

[0026] As a further solution of the present invention, the user rights management module provides user identity authentication, rights management and behavior monitoring, including:

[0027] User identity authentication ensures the uniqueness of each user's identity and the security of operations. The user rights management module first allows users to provide basic information (such as username, password, email address, etc.) for registration through the user registration mechanism, and supports OAuth third-party authentication to enhance the convenience and security of registration. The registration process includes preliminary verification of user information, such as sending a verification email or SMS verification code to ensure the authenticity of the user's identity. When the user logs in, the system uses a hash algorithm to encrypt and store the user's password, and further enhances security through two-factor authentication. After entering the password, the user must obtain and enter a one-time verification code via mobile phone or email to complete the login. The password management function allows users to change their password at any time and provides a password retrieval function. The retrieval process is verified through the email or mobile phone number provided by the user when registering. In addition, the user rights management module limits the number of failed logins. After multiple consecutive failures, the account will be temporarily locked to prevent brute force attacks, thereby fully protecting the security of the user's identity and the overall security of the system.

[0028] Permission management assigns specific roles to each user to manage permissions. Different roles correspond to different sets of operational permissions. The User Permission Management module predefines multiple roles, such as general user, doctor, and administrator, covering a wide range of operational permissions, from basic data viewing to advanced data management. Furthermore, administrators can create custom roles and permissions based on specific needs. Flexible permission assignment methods are provided, allowing for both bulk management by user group and setting exclusive permissions for individual users. Permission assignment and changes follow an automatic inheritance mechanism. When role permissions are modified, the permissions of all associated users are updated simultaneously, ensuring timely and consistent management. To prevent abuse of permissions, the User Permission Management module implements a strict permission isolation mechanism, preventing low-privilege users from unauthorized access to high-privilege data. Furthermore, the User Permission Management module includes detailed operation logging and real-time abnormal behavior monitoring. Every permission change and operation is recorded and audited, allowing administrators to track and analyze user behavior at any time. Combined with data encryption, permission isolation, and multi-layered security protection, the User Permission Management module ensures the security and efficiency of permission management. Behavior monitoring keeps a detailed record of every user operation, including the time of operation, user identity information, and the specific object of the operation.

[0029] As a further solution of the present invention, the security protection module adopts end-to-end encryption to ensure that the data is always encrypted from the sending end to the receiving end, including:

[0030] Prostate ultrasound imaging data is encrypted using an AES algorithm at the transmitter. This encryption process converts the original plaintext data into a seemingly random and meaningless ciphertext. The encryption key is known only to the sender and receiver, and is securely shared and negotiated. When the encrypted data is transmitted over the network, passing through routers and servers, these intermediate nodes lack the decryption key and can only process the encrypted ciphertext data, unable to understand its true meaning. Upon receiving the encrypted data, the receiver uses the pre-shared or negotiated key to decrypt the ciphertext, thus restoring the original plaintext data. Throughout the entire process, data encryption and decryption operations are performed only at the sender and receiver; no intermediate link can obtain the plaintext content of the data.

[0031] As a further solution of the present invention, a method for collaborative cloud-based medical data management is provided, wherein the steps are as follows:

[0032] Step S1, collecting prostate ultrasound image data from different medical institutions and doctors' personal devices through the image data acquisition module and uploading it to the cloud storage module;

[0033] Step S2: providing user authentication, rights management and behavior recording functions through the user rights management module;

[0034] In step S3, the collaborative management module establishes a real-time collaborative operation queue model, stores the prostate ultrasound image data in the cloud server, and allows multiple users to access the prostate ultrasound image data in the cloud. By using the gradient descent algorithm, the optimal solution of the objective function of the real-time collaborative operation queue model is obtained.

[0035] The technical effects and advantages of the cloud-based collaborative medical data management system and its use method of the present invention are as follows:

[0036] The present invention uses an image data acquisition module to collect prostate ultrasound image data from personal devices of different medical institutions and doctors and upload it to a cloud storage module. The user authority management module provides user authentication, authority management, and behavior recording functions. The security protection module uses end-to-end encryption to ensure that the data is always encrypted throughout the entire process from the sender to the receiver. The collaborative management module optimizes the objective function of the established real-time collaborative operation queue model to ensure data security while achieving efficient cloud-based collaborative processing, allowing multiple users to operate concurrently on shared data. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 This is a schematic diagram of the structure of a cloud-based collaborative medical data management system of the present invention;

[0038] Figure 2This is a flow chart of a method for cloud-based collaborative medical data management according to the present invention. DETAILED DESCRIPTION

[0039] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0040] Example 1

[0041] The present invention provides a cloud-based collaborative medical data management system, comprising an image data acquisition module, a cloud storage module, a data processing module, a user authority management module, a security protection module and a collaborative management module. The image data acquisition module collects prostate ultrasound image data from different medical institutions and doctors' personal devices and uploads it to the cloud storage module; the data processing module comprises a data analysis unit and an image processing function unit, supporting parallel processing of large-scale data; the user authority management module provides user identity authentication, authority management and behavior recording functions; the security protection module adopts end-to-end encryption to ensure that the data is always in an encrypted state throughout the entire process from the sending end to the receiving end; the collaborative management module optimizes the objective function of the established real-time collaborative operation queue model for multiple users to operate concurrently on shared data, the image data acquisition module is connected to the cloud storage module, the data processing module is connected to the cloud storage module, the user authority management module is connected to the data processing module, the security protection module is connected to the user authority management module, and the collaborative management module is connected to the security protection module.

[0042] Furthermore, the image data acquisition module collects prostate ultrasound image data from medical institutions and doctors' personal devices and uploads it to the cloud. This includes: the image data acquisition module connects with the medical information systems of each institution, adopts the HL7 international medical information exchange standard, and stably obtains prostate ultrasound image data from the servers of each medical institution; for doctors' personal devices, the image data acquisition module provides a dedicated application that supports Windows, iOS, and Android operating systems. Doctors install the dedicated application and upload the prostate ultrasound image data stored on their personal devices to the image data acquisition module;

[0043] The collected prostate ultrasound image data has three formats: DICOM, JPEG and PNG. The image data acquisition module converts the three different formats of prostate ultrasound image data into a unified format, and uses the OpenCV open source image processing library to convert the prostate ultrasound image from JPEG format and PNG format to DICOM format. The prostate ultrasound image data converted into the unified DICOM format is set to be automatically uploaded to the cloud at 2 am every day. During the upload process, the upload progress is displayed in real time, including the amount of uploaded data and the remaining time information. The user can understand the progress of the upload through the progress bar. After the upload is completed, the image data acquisition module uses a hash algorithm to perform integrity verification on the uploaded data to ensure that the data is not lost or damaged during transmission.

[0044] Furthermore, the cloud storage module provides data storage capabilities and is composed of a distributed database unit and a cloud storage technology unit, including:

[0045] The distributed database unit achieves high data availability and fault tolerance by distributing prostate imaging data across multiple servers and nodes. The distributed database unit also includes data sharding, data replication, data consistency, availability, and partition tolerance. Data sharding divides data into multiple parts based on patient ID, with each part stored on a different node. To improve data fault tolerance and availability, data is typically replicated to multiple nodes. Distributed databases require a balance between consistency, availability, and partition tolerance.

[0046] The cloud storage technology unit provides elastic, scalable, and highly available storage solutions. Cloud storage is combined with object storage services to manage large-scale unstructured data, such as prostate ultrasound images and patient information. The cloud storage technology unit stores data in the form of objects in a logically unified storage pool and uses redundancy technology to ensure high data availability.

[0047] Furthermore, the data processing module includes a data analysis unit and an image processing functional unit, supporting parallel processing of large-scale data, including:

[0048] The data analysis unit performs multi-dimensional quantitative analysis on the prostate ultrasound images, reduces the noise of the prostate ultrasound images through median filtering, performs equalization on the grayscale histogram of the prostate ultrasound images to enhance the image contrast, uses the Canny edge detection algorithm to extract the prostate boundary information in the prostate ultrasound images, and uses the deep learning-based segmentation algorithm U-Net to identify and segment nodules in the prostate ultrasound images and annotate them. By performing three-dimensional reconstruction on multiple frames of prostate ultrasound images, the three-dimensional structure of the prostate is reconstructed to provide intuitive visualization information and establish a multi-frame ultrasound three-dimensional pixel reconstruction model. The formula of the multi-frame ultrasound three-dimensional pixel reconstruction model is:

[0049] ,

[0050] Where: is the horizontal position of the prostate ultrasound image, is the vertical position of the prostate ultrasound image, is the depth coordinate in three-dimensional space, representing the The position of the prostate ultrasound image frame in three-dimensional space, In three-dimensional space, the coordinates Determine a three-dimensional pixel value, After image registration processing The pixel value of the frame prostate ultrasound image on the image plane, where , For the The pixel value of the frame prostate ultrasound image on the image plane, is a two-dimensional affine transformation matrix, For the Pixel values ​​in the frame 2D prostate ultrasound image;

[0051] The image processing functional unit uses a multi-core processor to decompose image processing tasks into multiple threads for parallel execution, thereby improving processing efficiency. Through stream processing technology, it processes prostate ultrasound image data in real time and provides processing results in a timely manner. It uses GPUs for large-scale matrix operations to accelerate the execution of image processing algorithms.

[0052] In this embodiment, in addition to the aforementioned large-scale matrix operations using the GPU, the GPU can also be used for convolution operations. A 3x3 convolution kernel is designed and uploaded from the host to the GPU. In order to enable the GPU to quickly access the kernel when performing convolution calculations, the convolution operation is started on the GPU. All threads work in parallel, and each thread independently calculates a part of the image. This process greatly speeds up the calculation. After the convolution operation is completed, the processing result is copied from the GPU memory back to the host memory.

[0053] Furthermore, the user rights management module provides user identity authentication, rights management, and behavior monitoring, including:

[0054] User identity authentication ensures the uniqueness of each user's identity and the security of operations. The user rights management module first allows users to provide basic information (such as username, password, email address, etc.) for registration through the user registration mechanism, and supports OAuth third-party authentication to enhance the convenience and security of registration. The registration process includes preliminary verification of user information, such as sending a verification email or SMS verification code to ensure the authenticity of the user's identity. When the user logs in, the system uses a hash algorithm to encrypt and store the user's password, and further enhances security through two-factor authentication. After entering the password, the user must obtain and enter a one-time verification code via mobile phone or email to complete the login. The password management function allows users to change their password at any time and provides a password retrieval function. The retrieval process is verified through the email or mobile phone number provided by the user when registering. In addition, the user rights management module limits the number of failed logins. After multiple consecutive failures, the account will be temporarily locked to prevent brute force attacks, thereby fully protecting the security of the user's identity and the overall security of the system.

[0055] Permission management assigns specific roles to each user to manage permissions. Different roles correspond to different sets of operation permissions. The user permission management module predefines multiple roles, such as ordinary users, doctors, and administrators, covering various types of operation permissions from basic data viewing to advanced data management. In addition, it allows administrators to create custom roles and permissions based on specific needs and provides flexible permission allocation methods. They can perform batch management by user groups or set exclusive permissions for individual users. The allocation and change of permissions follow an automatic inheritance mechanism. When role permissions are modified, the permissions of all associated users will be updated synchronously to ensure timely and consistent management. To prevent permission abuse, the user permission management module implements a strict permission isolation mechanism. Low-privilege users cannot unauthorize access to high-privilege data. At the same time, the user permission management module includes detailed operation log records and real-time abnormal behavior monitoring functions. Every permission change and operation behavior will be recorded and audited. Administrators can track and analyze user behavior at any time. Combined with data encryption, permission isolation and multi-level security protection, the user permission management module ensures the security and efficiency of permission management.

[0056] Behavior monitoring keeps detailed records of every user operation, including the time of the operation, user identity information, and the specific object of the operation. Rules and thresholds are set to detect abnormal behavior. If the same user fails to log in more than 10 times in a minute, it is considered abnormal behavior. Editing the same data file more than 20 times within an hour is considered abnormal behavior. Normal working hours are set from 9:00 to 20:00, and more than three operations outside normal working hours are considered abnormal behavior. The amount of data a single user can download in a day is limited to 10GB, and downloads exceeding 10GB are considered abnormal behavior. When abnormal behavior is detected, the user rights management module will immediately issue an alarm and notify the administrator, so that the administrator is immediately aware of the user's operation dynamics. Behavior monitoring also regularly analyzes user behavior data and generates behavior reports. These reports help administrators understand user usage habits, operation patterns, and potential risks. Through in-depth analysis of behavior data, loopholes and deficiencies in the user rights management module's permission settings and operation processes are discovered, providing a basis for further optimization and improvement.

[0057] Furthermore, the security protection module uses end-to-end encryption to ensure that data is always encrypted from the sender to the receiver, including:

[0058] Prostate ultrasound imaging data is encrypted using an AES algorithm at the transmitter. This encryption process converts the original plaintext data into a seemingly random and meaningless ciphertext. The encryption key is known only to the sender and receiver, and is securely shared and negotiated. When the encrypted data is transmitted over the network, passing through routers and servers, these intermediate nodes lack the decryption key and can only process the encrypted ciphertext data, unable to understand its true meaning. Upon receiving the encrypted data, the receiver uses the pre-shared or negotiated key to decrypt the ciphertext, thus restoring the original plaintext data. Throughout the entire process, data encryption and decryption operations are performed only at the sender and receiver; no intermediate link can obtain the plaintext content of the data.

[0059] Furthermore, the collaborative management module optimizes the objective function of the established real-time collaborative operation queue model using a gradient descent algorithm, enabling efficient cloud-based collaborative processing in a cloud-based collaborative medical data management system, allowing multiple users to concurrently operate on shared data, including:

[0060] The collaborative management module establishes a real-time collaborative operation queue model to store prostate ultrasound image data in the cloud server, allowing multiple users to access the prostate ultrasound image data in the cloud. The formula of the real-time collaborative operation queue model is:

[0061] ,

[0062] Where: For real-time collaborative operation queue model, is the total number of data sharing and real-time collaborative operations, Index variables for data sharing and real-time collaborative operations, is the average response time, is the weight associated with the dependency between data sharing and real-time collaborative operations, For the data sharing and real-time collaborative operation response time, is the throughput, The total time required to complete all data sharing and real-time collaborative operations, For the The contribution factors of data sharing and real-time collaborative operations to model throughput, is the memory resource utilization, is the total number of memory resource types, is the memory resource type, Memory resource type The total amount, For the Data sharing and real-time collaborative operation on memory resources The specific demand, For collaborative operation queue;

[0063] The objective function of the real-time collaborative operation queue model is:

[0064] ,

[0065] Where: is the weight coefficient of average response time, is the throughput weight coefficient, is the weight coefficient of memory resource utilization. In order to optimize the objective function, the gradient descent algorithm is used. The objective function The partial derivative with respect to the average response time is , the objective function The partial derivative with respect to throughput is , the objective function The partial derivative of memory resource utilization is , the formula for updating variables according to the gradient descent method is:

[0066]

[0067]

[0068] ,

[0069] Where: is the learning rate, set to 0.01, No. data sharing and real-time collaborative operation response time, After the update data sharing and real-time collaborative operation response time, The total time required to complete all data sharing and real-time collaborative operations, The total time required to complete all data sharing and real-time collaboration operations after the update. For the Data sharing and real-time collaborative operation on memory resources The specific demand, After the update Data sharing and real-time collaborative operation on memory resources The specific demand, repeat the above steps, constantly update the variables, and detect the objective function The rate of change in each iteration is dynamically determined to stop the iteration when the objective function The rate of change is less than for 10 consecutive iterations , stop iteration, objective function The optimal solution was reached.

[0070] The present invention provides a method for cloud-based collaborative medical data management, which comprises the following steps:

[0071] Step S1, collecting prostate ultrasound image data from different medical institutions and doctors' personal devices through the image data acquisition module and uploading it to the cloud storage module;

[0072] Step S2: providing user authentication, rights management and behavior recording functions through the user rights management module;

[0073] In step S3, the collaborative management module establishes a real-time collaborative operation queue model, stores the prostate ultrasound image data in the cloud server, and allows multiple users to access the prostate ultrasound image data in the cloud. By using the gradient descent algorithm, the optimal solution of the objective function of the real-time collaborative operation queue model is obtained.

[0074] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0075] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A cloud-based collaborative medical data management system, comprising an image data acquisition module, a cloud storage module, a data processing module, a user authority management module, a security protection module, and a collaborative management module, characterized in that: The image data acquisition module collects prostate ultrasound image data and uploads it to the cloud storage module; the data processing module includes a data analysis unit and an image processing functional unit, supporting parallel data processing; the user rights management module provides user authentication, rights management, and behavior recording functions; the security protection module uses end-to-end encryption to ensure that data is always encrypted throughout the entire process from the sender to the receiver; the collaborative management module uses a gradient descent algorithm to obtain the optimal solution to the objective function of the real-time collaborative operation queue model, realizing efficient cloud-based collaborative processing in the cloud-based collaborative medical data management system. The formula for the objective function of the real-time collaborative operation queue model is: ， Where: The objective function of the real-time collaborative operation queue model is is the weight coefficient of average response time, is the throughput weight coefficient, is the weight coefficient of memory resource utilization, is the total number of data sharing and real-time collaborative operations, Index variables for data sharing and real-time collaborative operations, is the average response time, is the weight associated with the dependency between data sharing and real-time collaborative operations, For the data sharing and real-time collaborative operation response time, is the throughput, The total time required to complete all data sharing and real-time collaborative operations, For the The contribution factors of data sharing and real-time collaborative operations to model throughput, is the memory resource utilization, is the total number of memory resource types, is the memory resource type, Memory resource type The total amount, For the Data sharing and real-time collaborative operation on memory resources The specific demand, A collaborative operation queue.

2. A cloud-based collaborative medical data management system according to claim 1, characterized in that ,Using the gradient descent algorithm, we can get the optimal solution of the objective function of the real-time collaborative operation queue model. The partial derivative with respect to the average response time is , the objective function The partial derivative with respect to throughput is , the objective function The partial derivative of memory resource utilization is , the formula for updating variables according to the gradient descent method is: ， Where: is the learning rate, set to 0.01, No. data sharing and real-time collaborative operation response time, After the update data sharing and real-time collaborative operation response time, The total time required to complete all data sharing and real-time collaborative operations, The total time required to complete all data sharing and real-time collaboration operations after the update. For the Data sharing and real-time collaborative operation on memory resources The specific demand, After the update Data sharing and real-time collaborative operation on memory resources The specific demand, repeat the above steps, constantly update the variables, and detect the objective function The rate of change in each iteration is dynamically determined to stop the iteration when the objective function The rate of change is less than for 10 consecutive iterations , stop the iteration, then the objective function Reach the optimal solution.

3. A cloud-based collaborative medical data management system according to claim 1, characterized in that , The image data acquisition module is connected to the cloud storage module, the data processing module is connected to the cloud storage module, the user authority management module is connected to the data processing module, the security protection module is connected to the user authority management module, and the collaborative management module is connected to the security protection module.

4. A cloud-based collaborative medical data management system according to claim 1, characterized in that: The data processing module includes a data analysis unit and an image processing unit, which supports parallel processing of large-scale data. The data analysis unit reconstructs the three-dimensional structure of the prostate by performing three-dimensional reconstruction on multiple frames of prostate ultrasound images, provides intuitive visualization information, and establishes a multi-frame ultrasound three-dimensional pixel reconstruction model. The formula of the multi-frame ultrasound three-dimensional pixel reconstruction model is: ， Where: is the horizontal position of the prostate ultrasound image, is the vertical position of the prostate ultrasound image, is the depth coordinate in three-dimensional space, representing the The position of the prostate ultrasound image frame in three-dimensional space, In three-dimensional space, the coordinates Determine a three-dimensional pixel value, After image registration processing The pixel value of the frame prostate ultrasound image on the image plane, where , For the The pixel value of the frame prostate ultrasound image on the image plane, is a two-dimensional affine transformation matrix, For the The image processing functional unit uses a multi-core processor to decompose the image processing task into multiple threads for parallel execution, thereby improving processing efficiency. Through stream processing technology, it processes the prostate ultrasound image data in real time and provides processing results in a timely manner. It uses GPU to perform large-scale matrix operations to accelerate the execution of image processing algorithms.

5. A method for using cloud-based collaborative medical data management, for implementing a cloud-based collaborative medical data management system according to any one of claims 1 to 4, characterized in that: Step S1, collecting prostate ultrasound image data through the image data acquisition module and uploading it to the cloud storage module; Step S2: providing user authentication, rights management and behavior recording functions through the user rights management module; In step S3, the collaborative management module establishes a real-time collaborative operation queue model, stores the prostate ultrasound image data in the cloud server, and enables multiple users to access the prostate ultrasound image data in the cloud. By using the gradient descent algorithm, the optimal solution of the objective function of the real-time collaborative operation queue model is obtained.

Images