Lossless compression method and system for medical image data

Abstract

The invention belongs to the technical field of data compression, and discloses a lossless compression method and system for medical image data. The method comprises the following steps: constructing a medical image data lossless compression engine in a cloud data center by using an image processing and deep learning algorithm according to a plurality of historical medical image data; using a data terminal to preprocess the collected real-time medical image data, and uploading the obtained preprocessed real-time medical image data to a cloud data center; and performing lossless compression on the preprocessed real-time medical image data received by the cloud data center by using a medical image data lossless compression engine to obtain real-time lossless compression data. The problems of low compression efficiency, high computing resource consumption and low compression quality in the prior art are solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of data compression, and particularly relates to a lossless compression method and system for medical image data. Background Art

[0002] Medical image data refers to the image materials obtained through imaging techniques during medical diagnosis and treatment. These data are usually generated by various medical imaging devices, such as X-ray films, computed tomography, magnetic resonance imaging, ultrasound examinations, or positron emission tomography. With the development of medical imaging technology, the amount of medical image data has increased rapidly, posing higher requirements for the storage and transmission of medical image data.

[0003] The existing medical image data compression technologies have the following defects:

[0004] 1) Low compression efficiency: The existing compression technologies often rely on traditional compression algorithms, such as image compression algorithms like JPEG 2000. These algorithms have low compression efficiency when dealing with high-resolution medical image data, resulting in a still relatively large volume of compressed data, which is not conducive to storage and transmission.

[0005] 2) High consumption of computing resources: The existing compression technologies require a large amount of computing resources during the compression process. Especially for multi-modal medical image data, the computational complexity is high, resulting in a long compression time, which is not conducive to real-time applications.

[0006] 3) Low compression quality: It is difficult for the existing compression technologies to find a balance between compression quality and compression speed, resulting in the compression quality of medical image data being difficult to meet the high-precision requirements of medical images. Summary of the Invention

[0007] In order to solve the problems of low compression efficiency, high consumption of computing resources, and low compression quality existing in the prior art, the purpose of the present invention is to provide a lossless compression method and system for medical image data.

[0008] The technical solution adopted by the present invention is as follows:

[0009] A lossless compression method for medical image data includes the following steps:

[0010] Based on a number of historical medical image data, use image processing and deep learning algorithms to construct a lossless compression engine for medical image data in the cloud data center;

[0011] Use a data terminal to preprocess the collected real-time medical image data and upload the preprocessed real-time medical image data to the cloud data center;

[0012] Use a lossless compression engine for medical image data to perform lossless compression on the preprocessed real-time medical image data received by the cloud data center, obtaining real-time lossless compression data.

[0013] Further, based on a number of historical medical image data, use image processing and deep learning algorithms to construct a lossless compression engine for medical image data in the cloud data center, including the following steps:

[0014] Based on the cloud data center, collect a number of historical medical image data and preprocess the number of historical medical image data to obtain a number of preprocessed historical medical image data;

[0015] According to a number of preprocessed historical medical image data, use image processing algorithms to construct a medical image data processing model and obtain a number of image-processed historical medical image data;

[0016] According to a number of preprocessed historical medical image data, use deep learning algorithms to construct a medical image data feature extraction model and obtain a number of historical medical image data features;

[0017] According to a number of historical medical image data features, use deep learning algorithms to construct a lossless compression model for medical image data;

[0018] Integrate the medical image data processing model, the medical image data feature extraction model, and the medical image data lossless compression model to obtain a lossless compression engine for medical image data.

[0019] Further, the medical image data processing model is constructed based on the SAM-Med2D-k-NN algorithm, and the medical image data processing model includes an image segmentation module constructed based on the SAM-Med2D algorithm and a graph structure construction module constructed based on the k-NN algorithm connected in sequence.

[0020] Further, the medical image data feature extraction model is constructed based on the GCN-Attention algorithm, and the medical image data feature extraction model includes a graph structure feature extraction module constructed based on the GCN algorithm and an attention weighting module constructed based on the Attention mechanism connected in sequence.

[0021] Further, the medical image data lossless compression model is constructed based on the DBN algorithm.

[0022] Further, use the lossless compression engine for medical image data to perform lossless compression on the preprocessed real-time medical image data received by the cloud data center, obtaining real-time lossless compression data, including the following steps:

[0023] Use a cloud data center to receive the preprocessed real-time medical image data uploaded by a data terminal, and input the preprocessed real-time medical image data into a medical image data lossless compression engine;

[0024] Use a medical image data processing model to process the preprocessed real-time medical image data to obtain real-time medical image data after image processing;

[0025] Use a medical image data feature extraction model to extract features from the real-time medical image data after image processing to obtain real-time medical image data features;

[0026] Use a medical image data lossless compression model to perform lossless compression on the real-time medical image data features to obtain real-time lossless compressed data.

[0027] Further, using a medical image data processing model to process the preprocessed real-time medical image data to obtain real-time medical image data after image processing includes the following steps:

[0028] Use the image segmentation module of the medical image data processing model to segment the preprocessed real-time medical image data to obtain several segmented images of the real-time medical image data;

[0029] Use the graph structure construction module of the medical image data processing model to construct real-time association relationships between each segmented image of the real-time medical image data and other segmented images of the real-time medical image data;

[0030] Use the segmented images of the real-time medical image data as nodes and the real-time association relationships as the edges of the nodes to construct a graph structure to obtain real-time medical image data after image processing in the form of a graph structure.

[0031] Further, using a medical image data feature extraction model to extract features from the real-time medical image data after image processing to obtain real-time medical image data features includes the following steps:

[0032] Use the graph structure feature extraction module of the medical image data feature extraction model to extract several real-time node features and several real-time edge features of the real-time medical image data after image processing;

[0033] Based on the several real-time node features and several real-time edge features, obtain the real-time graph structure data features of the real-time medical image data after image processing;

[0034] Use the attention weighting module of the medical image data feature extraction model to weight the real-time graph structure data features according to a preset attention weight value to obtain real-time medical image data features.

[0035] Further, using a lossless compression model for medical image data, perform lossless compression on the characteristics of real-time medical image data to obtain real-time lossless compression data, including the following steps:

[0036] Use the lossless compression model for medical image data to divide the characteristics of real-time medical image data into sub-blocks to obtain a number of real-time sub-blocks;

[0037] Encode each real-time sub-block to obtain a number of real-time compression codewords, and merge the number of real-time compression codewords to obtain initial real-time lossless compression data;

[0038] Perform post-processing on the initial real-time lossless compression data to obtain the final real-time lossless compression data.

[0039] A lossless compression system for medical image data, used to implement the lossless compression method. The system is deployed in a cloud data center, and the cloud data center is communicatively connected to a number of data terminals. The system includes an engine construction unit and a lossless compression unit connected in sequence;

[0040] The lossless compression unit is provided with a lossless compression engine for medical image data, and the lossless compression unit includes a medical image data processing sub-unit, a medical image data feature extraction sub-unit, and a medical image data lossless compression sub-unit connected in sequence.

[0041] The beneficial effects of the present invention are as follows:

[0042] The lossless compression method and system for medical image data provided by the present invention construct a lossless compression engine for medical image data, which can more effectively learn the high-level characteristics of medical image data, thereby achieving a higher compression ratio, significantly reducing the data volume, and improving the compression, storage, and transmission efficiency; constructing a lossless compression engine for medical image data in the cloud data center can utilize the powerful computing power of cloud computing to distributively process image data, reduce the computing burden of a single device, and improve the compression speed; the lossless compression engine for medical image data can achieve fast compression while maintaining high compression quality on the premise of ensuring the losslessness of image data through a series of image processing, feature extraction, and lossless compression; it can adapt to different types of medical image data without separately designing a compression algorithm for each image type; the generated real-time lossless compression data provides an efficient solution for the long-term storage, remote transmission, and online analysis of medical image data, promoting the development of medical informatization.

[0043] Other beneficial effects of the present invention will be further described in the specific implementation manner. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 is a flowchart of the lossless compression method for medical image data in the present invention.

[0045] Figure 2 It is the structural block diagram of the lossless compression system for medical image data in the present invention. Specific implementation manners

[0046] The present invention will be further explained below in conjunction with the accompanying drawings and specific embodiments.

[0047] Embodiment 1:

[0048] As Figure 1 shown, this embodiment provides a lossless compression method for medical image data, including the following steps:

[0049] S1: According to a number of historical medical image data, use image processing and deep learning algorithms to construct a lossless compression engine for medical image data in the cloud data center, including the following steps:

[0050] S1-1: Based on the cloud data center, collect a number of historical medical image data, and preprocess the number of historical medical image data to obtain a number of preprocessed historical medical image data;

[0051] The preprocessing includes denoising, image enhancement, and size normalization performed in sequence to improve the quality of medical image data and provide data support for subsequent model construction;

[0052] S1-2: According to a number of preprocessed historical medical image data, use image processing algorithms to construct a medical image data processing model, and obtain a number of processed historical medical image data;

[0053] The medical image data processing model is constructed based on the Segment Anything Model-SAM Medicine 2Dimensionality Med2D-k-Nearest Neighbors (k-NN) algorithm, and the medical image data processing model includes an image segmentation module constructed based on the SAM-Med2D algorithm and a graph structure construction module constructed based on the k-NN algorithm connected in sequence;

[0054] S1-3: According to a number of preprocessed historical medical image data, use deep learning algorithms to construct a medical image data feature extraction model, and obtain a number of historical medical image data features;

[0055] The medical image data feature extraction model is constructed based on the Graph Convolutional Network (GCN)-Attention algorithm, and the medical image data feature extraction model includes a graph structure feature extraction module constructed based on the GCN algorithm and an attention weighting module constructed based on the Attention mechanism, which are connected in sequence;

[0056] S1-4: According to several historical medical image data features, use a deep learning algorithm to construct a lossless compression model for medical image data;

[0057] The lossless compression model for medical image data is constructed based on the Deep Belief Nets (DBN) algorithm;

[0058] S1-5: Integrate the medical image data processing model, the medical image data feature extraction model, and the lossless compression model for medical image data to obtain a lossless compression engine for medical image data;

[0059] S2: Use a data terminal to preprocess the collected real-time medical image data and upload the preprocessed real-time medical image data to the cloud data center;

[0060] S3: Use the lossless compression engine for medical image data to perform lossless compression on the preprocessed real-time medical image data received by the cloud data center to obtain real-time lossless compressed data, including the following steps:

[0061] S3-1: Use the cloud data center to receive the preprocessed real-time medical image data uploaded by the data terminal and input the preprocessed real-time medical image data into the lossless compression engine for medical image data;

[0062] S3-2: Use the medical image data processing model to perform medical image data processing on the preprocessed real-time medical image data to obtain post-image-processing real-time medical image data, including the following steps:

[0063] S3-2-1: Use the image segmentation module of the medical image data processing model to perform image segmentation on the preprocessed real-time medical image data to obtain several segmented images of real-time medical image data;

[0064] The image segmentation module based on the SAM-Med2D algorithm is used to segment the preprocessed real-time medical image data. SAM-Med2D is a semantic segmentation algorithm for medical images. It can identify different structures and tissues in the image and segment them into different regions. The output is several segmented images, and each segmented image represents a specific region or structure in the original image. This segmentation helps subsequent analysis and processing because it can highlight the regions of interest and provide more detailed feature descriptions for each region;

[0065] S3-2-2: Use the graph structure construction module of the medical image data processing model to construct real-time association relationships between each real-time medical image data segmentation image and other real-time medical image data segmentation images;

[0066] The k-NN algorithm is a simple machine learning algorithm. It determines the similarity between data points by comparing feature vectors. The purpose is to establish connections between segmented images and identify which segmented images are close to each other in the feature space. These association relationships will then be used as the edges of the graph structure to form a graph structure network, where each node represents a segmented image;

[0067] S3-2-3: Use the real-time medical image data segmentation images as nodes and the real-time association relationships as the edges of the nodes to represent the spatial relationships and feature similarities in the medical image data, and perform graph structure construction to obtain the processed real-time medical image data of the graph structure, providing a higher-level structured representation for the medical image data;

[0068] Through segmentation and graph structure construction, the original image data is converted into a higher-level representation. This representation not only retains the detailed information of the image but also reveals the mutual relationships between different regions. The medical image data processing model is of great significance for improving the processing efficiency of medical image data, enhancing the depth and accuracy of data analysis, and optimizing data storage and transmission;

[0069] S3-3: Use the medical image data feature extraction model to extract the features of the processed real-time medical image data to obtain the real-time medical image data features, including the following steps:

[0070] S3-3-1: Use the graph structure feature extraction module of the medical image data feature extraction model to extract several real-time node features and several real-time edge features of the processed real-time medical image data;

[0071] The graph structure feature extraction module based on the GCN algorithm is used to extract the features of nodes and edges from the real-time medical image data after image processing. GCN is a deep learning model specifically for graph data. It can perform convolutional operations on the graph structure to learn the feature representations of nodes. The purpose is to extract useful features from each segmented image (node) and their associations (edges). These features may include texture, shape, edge information, etc., which are crucial for understanding the image content;

[0072] S3-3-2: Obtain the real-time graph structure data features of the real-time medical image data after image processing according to a number of real-time node features and a number of real-time edge features;

[0073] The node features and edge features obtained by the graph structure feature extraction module are integrated to form a complete representation of the graph structure data features. This representation captures the spatial relationships and local features in the medical image data, resulting in a richer and more comprehensive data feature representation. It not only contains the information of individual nodes but also the interaction information between nodes. This feature representation is very important for subsequent analysis and tasks because it provides a deeper understanding of the image;

[0074] S3-3-3: Use the attention weighting module of the medical image data feature extraction model to weight the real-time graph structure data features according to the preset attention weight values to obtain the real-time medical image data features;

[0075] The Attention mechanism can automatically assign different weights to different parts of the graph, emphasizing important features and ignoring unimportant parts. Through the Attention mechanism, the model can learn which features are most important for the interpretation of medical image data and assign higher weights to these features. The weighted feature representation highlights the key information more prominently, which helps to improve the performance of subsequent tasks. Moreover, the Attention mechanism enables the model to adapt to different data distributions and task requirements, improving the generalization ability of the model;

[0076] The GCN-Attention algorithm of the medical image data feature extraction model combines the advantages of the graph convolutional network and the attention mechanism, providing an efficient feature extraction method. It uses GCN to capture the complex relationships in the graph structure and highlights important features through the attention mechanism. Functionally, it can generate more accurate and meaningful feature representations, which are crucial for applications such as medical image analysis, diagnosis, and treatment planning. In this way, medical image data can be better understood and utilized, thus providing support for clinical decision-making;

[0077] S3-4: Use the lossless compression model for medical image data to perform lossless compression on the features of real-time medical image data, obtaining real-time lossless compression data, including the following steps:

[0078] S3-4-1: Use the lossless compression model for medical image data to divide the features of real-time medical image data into sub-blocks, obtaining a number of real-time sub-blocks;

[0079] The features of medical image data are divided into multiple sub-blocks. The purpose is to decompose the entire data set into smaller and more manageable parts, which can improve the efficiency and flexibility of compression; through sub-block division, the model can perform more refined coding for each small block, thereby optimizing the compression ratio and compression speed. In addition, sub-block division helps to maintain the local structure of the data, which is crucial for lossless compression;

[0080] S3-4-2: Encode each real-time sub-block to obtain a number of real-time compression codewords, and merge the number of real-time compression codewords to obtain the initial real-time lossless compression data;

[0081] DBN is a deep learning model composed of multiple restricted Boltzmann machine (RBM) layers, which can learn the probability distribution of data. The encoding process involves converting the sub-blocks into a set of compression codewords, which are compact representations of the original data; the DBN model can learn an effective representation of the data, thereby removing redundant information during the encoding process and achieving efficient compression. Since it is lossless compression, the encoding process of the DBN model can ensure that all information is retained so that the original data can be fully restored during decompression;

[0082] S3-4-3: Post-process the initial real-time lossless compression data to obtain the final real-time lossless compression data;

[0083] Through post-processing, the data can be further compressed, reducing the bandwidth required for storage and transmission, ensuring that the compressed data will not be damaged due to errors during transmission or storage, and maintaining the reliability of the data. The post-processing step provides the necessary metadata for the decompression process, ensuring that the decompression can accurately restore the original data;

[0084] The function of the lossless compression model for medical image data is to use the data features learned by the DBN algorithm to achieve efficient lossless data compression through sub-block division, encoding, and post-processing steps; in principle, the DBN algorithm can discover the statistical laws in the data, thereby removing redundant information while maintaining the integrity of the data; this method not only improves the efficiency of data processing but also reduces the storage and transmission costs of medical image data, which is of great significance for the performance and reliability of medical information systems.

[0085] Example 2:

[0086] As Figure 2 shown, this embodiment provides a lossless compression system for medical image data to implement a lossless compression method. The system is deployed in a cloud data center, which is communicatively connected to a plurality of data terminals. The system includes an engine construction unit and a lossless compression unit connected in sequence;

[0087] The lossless compression unit is provided with a lossless compression engine for medical image data, and the lossless compression unit includes a medical image data processing subunit, a medical image data feature extraction subunit, and a medical image data lossless compression subunit connected in sequence;

[0088] The data terminal is used to collect real-time medical image data, preprocess the collected real-time medical image data, and upload the preprocessed real-time medical image data to the cloud data center;

[0089] The engine construction unit is used to construct a lossless compression engine for medical image data in the cloud data center according to a plurality of historical medical image data by using image processing and deep learning algorithms;

[0090] The lossless compression unit is used to perform lossless compression on the preprocessed real-time medical image data received by the cloud data center by using the lossless compression engine for medical image data to obtain real-time lossless compression data;

[0091] The medical image data processing subunit is used to process the preprocessed real-time medical image data by using a medical image data processing model to obtain real-time medical image data after image processing;

[0092] The medical image data feature extraction subunit is used to extract medical image data features from the real-time medical image data after image processing by using a medical image data feature extraction model to obtain real-time medical image data features;

[0093] The medical image data lossless compression subunit is used to perform lossless compression on the real-time medical image data features by using a medical image data lossless compression model to obtain real-time lossless compression data.

[0094] A lossless compression method and system for medical image data, and a constructed lossless compression engine for medical image data can more effectively learn high-level features of medical image data, thereby achieving a higher compression ratio, significantly reducing the data volume, and improving the compression, storage, and transmission efficiency; constructing a lossless compression engine for medical image data in a cloud data center can utilize the powerful computing power of cloud computing to process image data distributively, reduce the computing burden on a single device, and improve the compression speed; the lossless compression engine for medical image data can achieve fast compression while maintaining high compression quality through a series of image processing, feature extraction, and lossless compression on the premise of ensuring the losslessness of image data; it can adapt to different types of medical image data without separately designing a compression algorithm for each image type; the generated real-time lossless compression data provides an efficient solution for the long-term storage, remote transmission, and online analysis of medical image data, promoting the development of medical informatization.

[0095] Those skilled in the art should understand that the embodiments of the present invention may be provided as a method, a system, or a computer program product. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0096] The present invention is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for realizing the specified functions in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0097] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device, and the instruction device realizes the specified functions in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0098] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus, so that a series of operation steps are performed on the computer or other programmable apparatus to generate a computer-implemented process, thereby providing instructions for implementing the process Figure 1 in one process or multiple processes and / or boxes Figure 1 steps for the functions specified in one box or multiple boxes.

[0099] Those of ordinary skill in the art can understand that all or part of the steps in implementing the above facts and methods can be completed by instructing relevant hardware through a program. The involved program or the described program can be stored in a computer-readable storage medium. When the program is executed, it includes the following steps: At this time, the corresponding method steps are introduced. The storage medium can be ROM / RAM, magnetic disk, optical disk, etc.

[0100] The specific embodiments described above further elaborate on the purpose, technical solutions, and beneficial effects of the present invention. It should be understood that the above are only specific embodiments of the present invention and are not used to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A lossless compression method for medical image data, characterized in that: The steps include: Based on some historical medical imaging data, we use image processing and deep learning algorithms to build a medical imaging data lossless compression engine in the cloud data center; Using a data terminal, pre-processing the collected real-time medical image data, and uploading the obtained pre-processed real-time medical image data to a cloud data center; A medical imaging data lossless compression engine is used to losslessly compress the pre-processed real-time medical imaging data received by the cloud data center to obtain real-time lossless compressed data.

2. A lossless compression method for medical image data according to claim 1, characterized in that: Based on some historical medical imaging data, we use image processing and deep learning algorithms to build a medical imaging data lossless compression engine in a cloud data center, including the following steps: Based on the cloud data center, a number of historical medical imaging data are collected, and the number of historical medical imaging data are pre-processed to obtain a number of pre-processed historical medical imaging data; According to a number of pre-processed historical medical image data, using an image processing algorithm, a medical image data processing model is constructed, and a number of image-processed historical medical image data are obtained; Based on a number of pre-processed historical medical imaging data, a medical imaging data feature extraction model is constructed using a deep learning algorithm, and a number of historical medical imaging data features are obtained; Based on several historical medical imaging data features, a deep learning algorithm is used to build a lossless compression model for medical imaging data; The medical image data processing model, the medical image data feature extraction model and the medical image data lossless compression model are integrated to obtain a medical image data lossless compression engine.

3. A lossless compression method for medical image data according to claim 2, characterized in that: The medical image data processing model is constructed based on the SAM-Med2D-k-NN algorithm, and the medical image data processing model includes an image segmentation module constructed based on the SAM-Med2D algorithm and a graph structure construction module constructed based on the k-NN algorithm, which are connected in sequence.

4. A lossless compression method for medical image data according to claim 3, characterized in that: The medical imaging data feature extraction model is constructed based on the GCN-Attention algorithm, and the medical imaging data feature extraction model includes a graph structure feature extraction module constructed based on the GCN algorithm and an attention weighting module constructed based on the Attention mechanism, which are connected in sequence.

5. A lossless compression method for medical image data according to claim 4, characterized in that: The medical image data lossless compression model is constructed based on the DBN algorithm.

6. A lossless compression method for medical image data according to claim 5, characterized in that: Using a medical image data lossless compression engine, losslessly compressing the pre-processed real-time medical image data received by the cloud data center to obtain real-time lossless compressed data includes the following steps: Using a cloud data center, receiving the pre-processed real-time medical imaging data uploaded by a data terminal, and inputting the pre-processed real-time medical imaging data into a medical imaging data lossless compression engine; Using a medical image data processing model, performing medical image data processing on the pre-processed real-time medical image data to obtain real-time medical image data after image processing; Using a medical image data feature extraction model, extracting medical image data features from real-time medical image data after image processing to obtain real-time medical image data features; The medical image data lossless compression model is used to perform lossless compression on the real-time medical image data features to obtain real-time lossless compressed data.

7. A method for lossless compression of medical image data according to claim 6, characterized in that: Using the medical image data processing model, performing medical image data processing on the pre-processed real-time medical image data to obtain the real-time medical image data after image processing, including the following steps: Using the image segmentation module of the medical image data processing model, image segmentation is performed on the pre-processed real-time medical image data to obtain a number of real-time medical image data segmentation images; Using a graph structure building module of a medical image data processing model, building a real-time association relationship between each real-time medical image data segmentation image and other real-time medical image data segmentation images; The real-time medical imaging data segmentation images are used as nodes, and the real-time association relationships are used as edges of the nodes to construct a graph structure, thereby obtaining real-time medical imaging data after image processing of the graph structure.

8. A lossless compression method for medical image data according to claim 7, characterized in that: Using the medical image data feature extraction model, the medical image data feature extraction is performed on the real-time medical image data after image processing to obtain the real-time medical image data features, including the following steps: Use the graph structure feature extraction module of the medical image data feature extraction model to extract several real-time node features and several real-time edge features of the real-time medical image data after image processing; According to a number of real-time node features and a number of real-time edge features, real-time graph structure data features of real-time medical imaging data after image processing are obtained; Using the attention weighting module of the medical imaging data feature extraction model, the real-time graph structure data features are weighted according to the preset attention weight values ​​to obtain real-time medical imaging data features.

9. A lossless compression method for medical image data according to claim 8, characterized in that: Using a medical image data lossless compression model, the medical image data lossless compression is performed on the real-time medical image data features to obtain real-time lossless compressed data, including the following steps: Using a medical imaging data lossless compression model, the real-time medical imaging data features are divided into sub-blocks to obtain a number of real-time sub-blocks; Encoding each real-time sub-block to obtain a plurality of real-time compressed codewords, and merging the plurality of real-time compressed codewords to obtain initial real-time lossless compressed data; The initial real-time lossless compressed data is post-processed to obtain the final real-time lossless compressed data.

10. A lossless compression system for medical image data, used to implement the lossless compression method according to any one of claims 1 to 9, characterized in that: The system is deployed in a cloud data center, the cloud data center is communicatively connected to a number of data terminals, and the system includes an engine construction unit and a lossless compression unit connected in sequence; The lossless compression unit is provided with a medical image data lossless compression engine, and the lossless compression unit comprises a medical image data processing subunit, a medical image data feature extraction subunit and a medical image data lossless compression subunit which are connected in sequence.

Images