Medical data classification and grading safety management method and device combining large model and block chain technology, electronic equipment and storage medium

Abstract

The invention discloses a medical data classification and grading safety management method and device combining a large model and a block chain technology, electronic equipment and a storage medium. The method comprises the following steps: S1, constructing a medical data classification and grading large model; s2, classifying and grading the medical data; s3, data uplink and secure storage; s4, intelligent contract dynamic authorization and access control; and S5, dynamically updating classification and grading specifications. Through deep fusion of a large model and a block chain technology, intelligent and secure medical data classification and level-to-level management is realized, the intelligent level and accuracy of data management are improved through a federal learning technology, and the response speed and adaptability of the model to specification standard updating are improved through technologies such as fine adjustment; the responsibility tracing problem of data management is solved through the block chain technology; in addition, the efficiency and safety of medical data management are remarkably improved, and comprehensive and reliable technical support is provided for data value release of the medical industry.

Description

Technical Field

[0001] The present invention relates to the field of blockchain and big model technology, and in particular, to a medical data classification and grading security management method, device, electronic device and storage medium that combine big model and blockchain technology. Background Art

[0002] Medical data classification and grading security management refers to a management system that divides medical data into different categories and levels based on its sensitivity, importance, and potential risks, and implements differentiated security protections. Its core goal is to balance data security and value utilization, ensuring patient privacy and the compliant flow of medical information. Classification is typically based on data sources (such as patient medical records and public health data) and attributes (such as personal identity information and health status data), while grading is based on the potential harm caused by data leakage or tampering, and is generally divided into three levels: core data, important data, and general data.

[0003] The purposes of the above management mechanisms include:

[0004] 1. Strengthen security protection: By clarifying the level of sensitive data (such as patient identity information and medical history), targeted encryption, access control and other measures are taken to reduce the risk of leakage;

[0005] 2. Meet compliance requirements: Implement laws and regulations such as the Data Security Law and the Personal Information Protection Law to ensure the legality of data processing;

[0006] 3. Promote data sharing: Open non-sensitive data (such as desensitized statistical information) within a controllable scope to support applications in scenarios such as medical research and medical insurance payment;

[0007] 4. Optimize data governance: Improve the data management efficiency of medical institutions through asset sorting and standard setting, laying the foundation for precision medicine and smart hospital construction.

[0008] At present, the main processes for classification and grading management of existing medical data are as follows:

[0009] X1. Establish data classification and grading rules: Build a scientific and systematic data classification and grading framework and rules based on relevant specifications and standards;

[0010] X2. Data asset review: Comprehensively review the organization's structured and unstructured data assets, create an original list of data assets, and clarify the basic information of the data assets and related parties;

[0011] X3. Data asset-data classification and grading mapping: Database tables, fields, data items, data files, and other items in the original data asset list are mapped one by one to the data asset units in the data classification and grading rules through data element associations, clarifying the classification and grading of data assets.

[0012] X4. Data Classification and Grading - Asset Unit List Review: Review and optimize the data classification and grading - data asset mapping results;

[0013] X5. Data asset basic attribute annotation: According to the data asset description requirements and data asset catalog search requirements, the basic attributes of the data asset unit should be annotated;

[0014] X6. Data asset catalog review: Review and optimize the annotation results of basic data asset attributes to ultimately form a data asset catalog;

[0015] X7. Dynamic update management: Based on data classification elements and changes that may have an impact, dynamic update management is carried out on data classification and grading rules, data classification and grading-asset unit lists, basic attribute annotation sets and data asset catalogs.

[0016] The above management process has the following deficiencies:

[0017] ① The complexity of classification and grading rules and the professional and diverse characteristics of medical data mean that each step of data classification and grading management requires a large amount of professional human resources, which is costly and time-consuming, making it difficult for small medical institutions to bear and implement;

[0018] ② Medical data classification and grading management methods that rely on manual operations or basic rule engines are prone to misjudgment or omission, affecting the accuracy and reliability of the final classification and grading results;

[0019] ③ When classification and grading standards change, the existing system has a slow update response speed and cannot adapt to new rules and requirements in a timely manner. This not only affects the timeliness of data processing but also increases compliance risks.

[0020] ④ Existing data management methods usually lack traceability for data classification and grading and their storage addresses, and the management responsibilities are vaguely defined. They are unable to effectively track the data change history and usage records, making it difficult to trace responsibilities and audit when violations such as data leakage and tampering occur, and are unable to meet the increasingly stringent compliance requirements of the medical industry. Summary of the Invention

[0021] On the one hand, this application provides a medical data classification and grading security management method that combines a big model with blockchain technology to solve the technical problems of existing medical data classification and grading security management, such as high cost, poor accuracy and reliability, compliance risks, difficulty in responsibility tracing and auditing, and inability to meet the increasingly stringent compliance requirements of the medical industry.

[0022] This application is implemented through the following scheme:

[0023] A medical data classification and grading security management method combining a large model with blockchain technology includes the following steps:

[0024] S1. Construction of a large model for medical data classification and grading. After pre-training, fine-tuning, and parameter optimization training of the large model using public medical data, a large model for medical data classification and grading is obtained.

[0025] S2. Classification and grading of medical data: Use the trained medical data classification and grading model to classify and grade each piece of medical data, map it one by one to the data classification and grading rules, and add corresponding classification and grading labels;

[0026] S3, data on-chain and secure storage: Classified and graded medical data is processed and uploaded to the chain, recorded in a distributed ledger, and the correctness and integrity of the data storage are verified through the blockchain consensus mechanism to ensure that the data cannot be tampered with;

[0027] S4. Dynamic authorization and access control of smart contracts. After receiving a user request, the blockchain system calls the relevant smart contract to verify the user's access rights and returns the verification result.

[0028] S5. Dynamic update of classification and grading specifications. When the classification and grading specifications of medical data change, the dynamic update process is triggered, the knowledge base is updated, and the updated data is re-uploaded to the blockchain system, user access rights are adjusted, and data privacy and security are guaranteed.

[0029] Furthermore, the step S1 specifically includes the steps of:

[0030] S11. Pre-train large models using a large amount of public medical data to equip them with professional general medical knowledge.

[0031] S12. Based on existing policies and regulations, formulate a classification and grading framework based on the basic nature, business attributes, and potential risks of medical data;

[0032] S13. Each medical institution uses data sampling techniques to collect representative small sample data sets from its own private databases, including representative diagnosis and treatment data and rare case data, and manually classifies and labels the collected data through multi-expert evaluation to construct a high-quality fine-tuned data set. The labeling experts all have extensive knowledge in the medical field and have received certain data labeling training. Labeling is only performed when more than half of the expert group reaches a consensus. If there is any disagreement, a discussion will be held, and if there is still a large disagreement after the discussion, the data set will be eliminated.

[0033] S14. Deploy the pre-trained large model locally in each medical institution and fine-tune the pre-trained large model based on the personalized data of each medical institution through low-rank adaptation technology (LoRA). Under the condition of freezing the original parameters of the pre-trained model, train it on the labeled dataset and obtain the matrix of parameter changes based on the gradient descent optimization of the cross-entropy loss function:

[0034]

[0035] Among them, y i is the true label, is the probability predicted by the model, n is the number of categories, and the fine-tuning method is applied to obtain the parameter change matrix (ΔW1,…,ΔW N );

[0036] S15, the parameter change matrix (Δw1, ..., ΔW N ) Use homomorphic encryption and differential privacy encryption technology to mask the gradient information and send the encrypted parameter change matrix to the central aggregation server;

[0037] S16. Use a method that combines spatial anomaly, behavioral anomaly, and amplitude anomaly multi-indicator detection to detect poisoning attacks;

[0038] S17. Calculate the contribution of each node based on the local model quality gain, the amount of data provided and the training intensity, and the number of abnormal labels;

[0039] S18. The central server performs weighted security aggregation based on the contribution of each institution. The weight change matrix of the institution with a large contribution occupies a larger weight in the training. The parameter change matrix after weighted addition is added to the parameter matrix of the pre-trained large model, and the accuracy change is recorded. When the accuracy change is less than the expected value, the model is considered to have converged and the training is terminated. Otherwise, the global model parameter change is encrypted and sent to each institution node. After updating the local model, return to step S13 until the model training converges.

[0040] Furthermore, the step S16 specifically includes the following steps:

[0041] S161. First, the parameter change matrices uploaded by each organization are mapped to a unified vector space and the cosine similarity between them is calculated according to the following formula:

[0042]

[0043] The parameter changes of malicious nodes are significantly lower than the similarity of most normal nodes, forming isolated clusters. A threshold θ is set. When the average cosine similarity of node i with other nodes is less than the set threshold, it is considered an abnormal user.

[0044] S162. Next, combine the parameter change matrix of each medical institution with the base large model, record the resulting performance changes, and mark the user as an abnormal user when the model performance degradation value exceeds the normal range.

[0045] S163. Finally, the norm of the parameter change matrix uploaded by all nodes is counted. When a node has malicious attack behavior, a large variation of ΔW will often appear in an attempt to control the model. If the norm of a node is much larger than the mean, it will be marked as an abnormal user; warning feedback will be given to the abnormal user, requiring the node to re-check the data label or re-fine-tune, and accumulate abnormal labels for it. Its ΔW will be downgraded during aggregation. When the cumulative number exceeds the preset threshold, the organization node will be permanently eliminated.

[0046] Furthermore, the step S17 specifically includes:

[0047] S171. Use accuracy, F1 score, ROC and other indicators to record the local gain of ΔW of each mechanism to the global model, denoted as G;

[0048] S172. Count the number of labeled samples and training rounds used for fine-tuning for each medical institution, denoted as Q, to avoid "free-riding" by nodes with extremely small data sizes and very few training rounds;

[0049] S173. Calculate the contribution of each medical institution based on the number of times F the node was identified as an abnormal node during training using the following formula:

[0050]

[0051] The weight parameters a, β, and γ can be adjusted dynamically as needed. A minimum protection threshold is set to avoid complete power loss for small medical institutions.

[0052] Furthermore, the step S2 specifically includes the steps of:

[0053] S21. Collect structured and unstructured medical data, including electronic health records, medical images, laboratory test results, physician diagnostic reports, and patient medical records, from multiple source systems of medical institutions, such as hospital information systems, laboratory information management systems, image archiving and communication systems, and perform data cleaning to remove noise and redundant information;

[0054] S22. Build a knowledge base based on the latest credible medical information, including the latest medical data classification and grading management standards, cutting-edge papers published in important medical academic journals, and authoritative research reports;

[0055] S23. Through search enhancement technology, relevant knowledge and rules of medical data to be classified and graded are retrieved from the knowledge base, and combined with prompt engineering technology to assist the large model to achieve more accurate data classification and grading;

[0056] S24. Use the trained large model to classify and grade each piece of medical data, map them one by one to the data classification and grading rules, and add corresponding classification and grading labels.

[0057] Furthermore, the step S3 specifically includes the steps of:

[0058] S31. Calculate hash values ​​for the classified and graded medical data to ensure the uniqueness and integrity of the data;

[0059] S32. Use asymmetric encryption algorithms to encrypt medical data to ensure data security during transmission and storage;

[0060] S33. Upload the encrypted medical data and its classification and grading labels to the blockchain network, record them in the distributed ledger, and verify the correctness and integrity of the data storage through the blockchain consensus mechanism to ensure that the data cannot be tampered with.

[0061] Furthermore, the step S4 specifically includes the steps of:

[0062] S41. The user packages the request including the request body, target data object, and the operation to be performed to generate a request R:

[0063] R←F(S, O, A),

[0064] Among them, R represents user request, S represents subject attributes (including user unique ID and permission level), O represents object attributes (including data category and level), and A represents operation attributes (including data operations such as read and write);

[0065] After signing with the private key, send the public key, certificate signature, and timestamp to the blockchain medical data management system:

[0066] B←X{PK X ,Sign(R,SK X ), T1},

[0067] Among them, B represents blockchain, X represents user, PK X Represents the public key of user X, Sign() represents the digital signature, SK X represents the private key of user X, T1,…T m Indicates a timestamp;

[0068] S42. After receiving the user request, the blockchain system uses its public key to parse the request and calls the policy management contract to automatically match the data object corresponding to the request:

[0069] B{Sign(R,SK X ), P(R), T2}→X,

[0070] Among them, P represents the policy management contract, which is responsible for automatically matching the corresponding data objects according to user requests and policy rules;

[0071] S43. Call the permission authentication contract to verify the user's access rights:

[0072] B{V(X), T3}→B,

[0073] V represents the permission verification contract, which is responsible for verifying whether the user has the permission to access a specific data object. If the verification is successful, the data retrieval contract is called to return the data corresponding to the request:

[0074] B{D(R), T4}→X,

[0075] Among them, D represents the data retrieval contract, which is responsible for retrieving and returning the data object requested by the user from the blockchain. If the verification fails, it means that the user request does not meet the policy information in the policy management contract, and a rejection message is returned:

[0076] B{Refused,T4}→X,

[0077] Refused indicates the plain text of the rejection information.

[0078] Furthermore, the step S5 specifically includes the steps of:

[0079] S51. When the medical data classification and grading specifications change, reselect and label the fine-tune data set according to the new medical data classification and grading specifications, and repeat step S1;

[0080] S52. Update the knowledge base to ensure the timeliness and accuracy of retrieval information, provide the latest and most reliable knowledge support for the classification and grading management of medical data, and ensure the efficiency and practicality of retrieval enhancement technology;

[0081] S53. Update the classification and grading labels of the medical data using the updated large-scale model classification and grading architecture, and re-upload the updated data to the blockchain system;

[0082] S54. Adjust user access rights according to the updated medical data classification and grading specifications to ensure data privacy and security.

[0083] On the other hand, the present application also provides a medical data classification and grading security management device that combines a large model with blockchain technology, including:

[0084] The module for building a large model for medical data classification and grading is used to build a large model for medical data classification and grading. After pre-training, fine-tuning, and parameter optimization training of the large model using public medical data, the large model for medical data classification and grading is obtained.

[0085] The medical data classification and grading processing module is used for medical data classification and grading. It uses the trained medical data classification and grading model to classify and grade each piece of medical data, maps it one by one to the data classification and grading rules, and adds corresponding classification and grading labels.

[0086] The data on-chain and secure storage module is used for data on-chain and secure storage. It processes and uploads classified and graded medical data to the chain, records it in a distributed ledger, and verifies the correctness and integrity of the data storage through the blockchain consensus mechanism to ensure that the data cannot be tampered with.

[0087] Smart contract dynamic authorization and access control module, used for dynamic authorization and access control of smart contracts. After the blockchain system receives a user request, it calls the relevant smart contract to verify the user's access rights and returns the verification result;

[0088] The classification and grading specification dynamic update module is used to dynamically update the classification and grading specifications. When the classification and grading specifications of medical data change, the dynamic update process is triggered, the knowledge base is updated, and the updated data is re-uploaded to the blockchain system, user access rights are adjusted, and data privacy and security are guaranteed.

[0089] On the other hand, the present application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the medical data classification and grading security management method combining the big model with blockchain technology are implemented.

[0090] On the other hand, the present application also provides a storage medium, which includes a stored program, and when the program is running, controls the device where the storage medium is located to execute the steps of the medical data classification and grading security management method that combines the big model and blockchain technology.

[0091] Compared with the existing technology, this application has the following beneficial effects:

[0092] This application, through the deep integration of big data models and blockchain technology, has constructed an intelligent and secure medical data classification and grading management system. First, using federated learning technology, this application enables the collaborative training of a large-scale medical data classification and grading model without leaving the domain of data from multiple institutions. This enables the precise classification and grading of multimodal medical data, such as text and images, significantly improving the intelligence and accuracy of data management. Furthermore, by combining fine-tuning and retrieval enhancement techniques, this application improves the model's responsiveness and adaptability to regulatory standards updates. Second, by leveraging blockchain's distributed ledger technology, asymmetric encryption, and hashing algorithms, this application ensures the secure storage and immutability of medical data, addressing accountability issues in data management and providing a highly trusted data management environment for medical institutions, patients, and regulatory authorities. Furthermore, this application utilizes smart contract technology to achieve automated authentication and access control. Through predefined rules and logic, this application ensures that only authorized users can access data at a specific level, reducing the risk of data leakage and misuse. Through the synergy of big data models and blockchain, this application significantly improves the efficiency and security of medical data management, providing comprehensive and reliable technical support for unlocking the value of data in the medical industry.

[0093] In addition to the above-described purposes, features and advantages, the present application also has other purposes, features and advantages. The present application will be further described in detail below with reference to the drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0094] The drawings that constitute a part of this application are used to provide further understanding of this application. The illustrative embodiments of this application and their descriptions are used to explain this application and do not constitute improper limitations on this application.

[0095] Figure 1 This is a flow chart of the medical data classification and grading security management method that combines a big model with blockchain technology in a preferred embodiment of the present application.

[0096] Figure 2 It is a schematic diagram of the process of constructing a large model for medical data classification and grading in the preferred embodiment of the present application.

[0097] Figure 3 This is a schematic diagram of the medical data classification and grading process based on large model technology in the preferred embodiment of the present application.

[0098] Figure 4 This is a schematic diagram of user data access control management based on blockchain smart contracts in this application.

[0099] Figure 5 This is a schematic diagram of the module of a medical data classification and grading security management device that combines a large model with blockchain technology in another preferred embodiment of the present application.

[0100] Figure 6 This is a schematic block diagram of an electronic device entity according to a preferred embodiment of the present application.

[0101] Figure 7 It is a diagram of the internal structure of a computer device according to a preferred embodiment of the present application. DETAILED DESCRIPTION

[0102] The embodiments of the present application are described in detail below with reference to the accompanying drawings, but the present application can be implemented in a variety of different ways defined and covered below.

[0103] Explanation of relevant terms:

[0104] Data classification refers to the process of distinguishing and categorizing data into major categories, medium categories, minor categories, and subcategories based on certain common attributes or characteristics of the data and following certain principles or methods, in order to facilitate the refined management and use of data.

[0105] Data classification is based on data classification. It divides data into different levels according to the degree of impact caused by data tampering, destruction, illegal acquisition or illegal use on national security, economic operation, social stability, public interests or the legitimate rights and interests of individuals and organizations, so as to distinguish the degree of restriction on the scope of data access, use or disclosure.

[0106] Big model technology is an advanced artificial intelligence method based on deep learning technology. Its core is to use deep learning neural network architectures containing billions or even hundreds of billions of parameters to simulate human cognitive abilities in natural language learning, thereby achieving efficient processing and analysis of complex patterns and relationships contained in the data. By using advanced algorithms such as attention mechanisms and combining pre-training with massive multi-source heterogeneous data, big models can deeply explore subtle features and potential patterns in the data, thereby achieving high-precision prediction, classification, generation and decision-making capabilities. In addition, big models also have powerful transfer learning capabilities and can quickly adapt to new tasks and fields. With its excellent data representation capabilities and generalization performance, big models have demonstrated significant advantages in multimodal data processing fields such as natural language processing, computer vision, and speech recognition.

[0107] Model fine-tuning is an efficient AI technique based on transfer learning. Its core approach is to leverage the general features learned by a pre-trained model on large-scale datasets and further train it on smaller datasets for specific tasks, enabling rapid adaptation to new tasks and improving model performance. Specifically, fine-tuning adjusts some or all of the parameters of a pre-trained model to enable it to capture the specific patterns and regularities of the target task while retaining the extensive knowledge learned during pre-training. The advantage of fine-tuning is that it significantly reduces the amount of data required for the target task, lowering training costs while improving the model's accuracy and robustness on new tasks. Furthermore, fine-tuning supports layered tuning strategies, such as fine-tuning only the last few layers or specific modules of a model, enabling rapid adaptation to task requirements while preserving general features. Due to its efficiency, flexibility, and wide applicability, fine-tuning has been widely used in fields such as natural language processing (e.g., text classification and question-answering systems), computer vision (e.g., image segmentation and object detection), and speech recognition (e.g., dialect recognition and speech sentiment analysis), becoming a crucial tool for rapid model deployment and performance optimization.

[0108] Retrieval enhancement technology is an advanced artificial intelligence method that combines information retrieval and generative models. It aims to enhance the generation ability and accuracy of the model by introducing external knowledge sources. The core idea is to dynamically retrieve task-related information during the generation process, and use the retrieval results as context input to assist the model in generating more accurate and relevant outputs. Specifically, retrieval enhancement technology usually includes two key components: a retrieval module and a generation module. The retrieval module extracts the most relevant information to the input query from a large-scale knowledge base or document collection through an efficient search algorithm; the generation module uses a pre-trained language model to combine the retrieved information with the original input to generate high-quality text output. This technology can not only significantly improve the performance of the model in knowledge-intensive tasks, but also effectively reduce the risk of the model generating false or irrelevant content. In addition, retrieval enhancement technology is highly flexible and scalable. It only needs to update the data of the retrieval part to adapt to the needs of different fields and tasks, reducing the knowledge update cost of the model.

[0109] Blockchain technology is a data management method based on distributed ledgers and cryptographic principles. Its core principle is to build a highly secure and transparent data storage and transmission system through decentralization, immutability, and traceability. A blockchain consists of a series of chronologically connected blocks, each containing a cryptographically verified set of transaction data. A consensus mechanism ensures consistency across all nodes in the network regarding the state of data. Blockchain's distributed architecture eliminates reliance on centralized authority, enhancing the system's resilience and fault tolerance. Its immutability also provides a reliable foundation for data auditing and traceability.

[0110] Smart contracts are automated, programmable agreements based on blockchain technology. Their core principle is to automatically execute contract terms when specific conditions are met through predefined rules and logic, without the need for third-party intervention. Smart contracts leverage the decentralized and tamper-proof nature of blockchain to ensure transparent, secure, and reliable contract execution. Key technologies include a Turing-complete programming language, a state machine model, and an event-driven mechanism. These technologies enable smart contracts to handle complex business logic and automatically update and verify data on the blockchain. The execution process of smart contracts is completely transparent, with all operation records permanently stored on the blockchain for audit and traceability by all parties. Furthermore, smart contracts support multi-party collaboration, automatically coordinating and executing agreements between multiple participants, reducing human intervention and trust costs.

[0111] Federated learning is a distributed machine learning framework based on privacy-preserving and secure encryption technologies. Its core principle is to enable collaborative learning by training models locally and sharing only model parameters or gradients, ensuring data remains local. During the federated learning training process, data from all participants remains local, minimizing the risk of data leakage. Model parameter updates are transmitted and aggregated through encrypted channels, ensuring effective model training and data privacy compliance. Federated learning technology enables collaborative updates of global models across multiple data sources while avoiding direct exposure of raw data. This allows for cross-institutional and cross-regional collaborative data modeling while protecting data sovereignty.

[0112] like Figure 1 As shown, the preferred embodiment of the present application provides a medical data classification and grading security management method combining a large model with blockchain technology, including the following steps:

[0113] S1. Construction of a large model for medical data classification and grading. After pre-training, fine-tuning, and parameter optimization training of the large model using public medical data, a large model for medical data classification and grading is obtained.

[0114] S2. Classification and grading of medical data: Use the trained medical data classification and grading model to classify and grade each piece of medical data, map it one by one to the data classification and grading rules, and add corresponding classification and grading labels;

[0115] S3, data on-chain and secure storage: Classified and graded medical data is processed and uploaded to the chain, recorded in a distributed ledger, and the correctness and integrity of the data storage are verified through the blockchain consensus mechanism to ensure that the data cannot be tampered with;

[0116] S4. Dynamic authorization and access control of smart contracts. After receiving a user request, the blockchain system calls the relevant smart contract to verify the user's access rights and returns the verification result.

[0117] S5. Dynamic update of classification and grading specifications. When the classification and grading specifications of medical data change, the dynamic update process is triggered, the knowledge base is updated, and the updated data is re-uploaded to the blockchain system, user access rights are adjusted, and data privacy and security are guaranteed.

[0118] To address the shortcomings of the existing technology, the technical solution adopted in this embodiment is mainly divided into three parts, namely, the large model data classification and grading sub-part, the blockchain sub-part, and the smart contract sub-part, among which:

[0119] Large model data classification and grading sub-section: used to realize the automation and intelligent process of medical data classification and grading, simplifying a large number of manual processing operations and ensuring the accuracy of classification and grading results.

[0120] Blockchain sub-part: used to store medical data and its classification and grading labels to ensure that the data cannot be tampered with, is safe and reliable, and that the responsibility for data operations is traceable and management is transparent.

[0121] Smart contract sub-part: includes policy management contract, permission verification contract and data retrieval contract. The policy management contract is mainly used to formulate and execute management policies, automatically assign corresponding operation permissions according to the attributes of users or roles, and the permission verification contract is used to compare the requested subject permissions with the permission restrictions for accessing data, and perform access control through the policies in the policy contract. The data retrieval contract is used to retrieve and return the corresponding data according to user requests.

[0122] This embodiment, through the deep integration of big data models and blockchain technology, builds an intelligent and secure medical data classification and grading management system. First, this embodiment uses federated learning technology to collaboratively train a large medical data classification and grading model without leaving the domain of data from multiple institutions. This enables precise classification and grading of multimodal medical data, such as text and images, significantly improving the intelligence and accuracy of data management. Furthermore, by combining fine-tuning and search enhancement techniques, the model's responsiveness and adaptability to regulatory standards updates are enhanced. Second, by leveraging blockchain's distributed ledger technology, asymmetric encryption, and hashing algorithms, this embodiment ensures the secure storage and immutability of medical data, addressing accountability issues in data management and providing a highly trusted data management environment for medical institutions, patients, and regulatory authorities. Furthermore, this embodiment utilizes smart contract technology to achieve automated identity authentication and access control. Through predefined rules and logic, this ensures that only authorized users can access data at specific levels, reducing the risk of data leakage and misuse. Through the synergistic effect of big data models and blockchain, this embodiment significantly improves the efficiency and security of medical data management, providing comprehensive and reliable technical support for unlocking the value of data in the medical industry.

[0123] Preferably, if Figure 2 As shown, the step S1 specifically includes the following steps:

[0124] S11. Pre-train large models using a large amount of public medical data to equip them with professional general medical knowledge.

[0125] S12. Based on existing policies and regulations, formulate a classification and grading framework based on the basic nature, business attributes, and potential risks of medical data;

[0126] S13. Each medical institution uses data sampling techniques to collect representative small sample data sets from its own private databases, including representative diagnosis and treatment data and rare case data, and manually classifies and labels the collected data through multi-expert evaluation to construct a high-quality fine-tuned data set. The labeling experts all have extensive knowledge in the medical field and have received certain data labeling training. Labeling is only performed when more than half of the expert group reaches a consensus. If there is any disagreement, a discussion will be held, and if there is still a large disagreement after the discussion, the data set will be eliminated.

[0127] S14. Deploy the pre-trained large model locally in each medical institution and fine-tune the pre-trained large model based on the personalized data of each medical institution through low-rank adaptation technology (LoRA). Under the condition of freezing the original parameters of the pre-trained model, train it on the labeled dataset and obtain the matrix of parameter changes based on the gradient descent optimization of the cross-entropy loss function:

[0128]

[0129] Among them, y i is the true label, is the probability predicted by the model, n is the number of categories, and the fine-tuning method is applied to obtain the parameter change matrix (ΔW1,…,ΔW N );

[0130] S15, the parameter change matrix (Δw1, ..., ΔW N ) Use homomorphic encryption and differential privacy encryption technology to mask the gradient information and send the encrypted parameter change matrix to the central aggregation server;

[0131] S16. Use a method that combines spatial anomaly, behavioral anomaly, and amplitude anomaly multi-indicator detection to detect poisoning attacks;

[0132] S17. Calculate the contribution of each node based on the local model quality gain, the amount of data provided and the training intensity, and the number of abnormal labels;

[0133] S18. The central server performs weighted security aggregation based on the contribution of each institution. The weight change matrix of the institution with a large contribution occupies a larger weight in the training. The parameter change matrix after weighted addition is added to the parameter matrix of the pre-trained large model, and the accuracy change is recorded. When the accuracy change is less than the expected value, the model is considered to have converged and the training is terminated. Otherwise, the global model parameter change is encrypted and sent to each institution node. After updating the local model, return to step S13 until the model training converges.

[0134] Specifically, the step S16 includes the following steps:

[0135] S161. First, the parameter change matrices uploaded by each organization are mapped to a unified vector space and the cosine similarity between them is calculated according to the following formula:

[0136]

[0137] The parameter changes of malicious nodes are significantly lower than the similarity of most normal nodes, forming isolated clusters. A threshold θ is set. When the average cosine similarity of node i with other nodes is less than the set threshold, it is considered an abnormal user.

[0138] S162. Next, combine the parameter change matrix of each medical institution with the base large model, record the resulting performance changes, and mark the user as an abnormal user when the model performance degradation value exceeds the normal range.

[0139] S163. Finally, the norm of the parameter change matrix uploaded by all nodes is counted. When a node has malicious attack behavior, a large variation of ΔW will often appear in an attempt to control the model. If the norm of a node is much larger than the mean, it will be marked as an abnormal user; warning feedback will be given to the abnormal user, requiring the node to re-check the data label or re-fine-tune, and accumulate abnormal labels for it. Its ΔW will be downgraded during aggregation. When the cumulative number exceeds the preset threshold, the organization node will be permanently eliminated.

[0140] Specifically, the step S17 includes:

[0141] S171. Use accuracy, F1 score, ROC and other indicators to record the local gain of ΔW of each mechanism to the global model, denoted as G;

[0142] S172. Count the number of labeled samples and training rounds used for fine-tuning for each medical institution, denoted as Q, to avoid "free-riding" by nodes with extremely small data sizes and very few training rounds;

[0143] S173. Calculate the contribution of each medical institution based on the number of times F the node was identified as an abnormal node during training using the following formula:

[0144]

[0145] The weight parameters a, β, and γ can be adjusted dynamically as needed. A minimum protection threshold is set to avoid complete power loss for small medical institutions.

[0146] Preferably, if Figure 3 As shown, the step S2 specifically includes the following steps:

[0147] S21. Collect structured and unstructured medical data, including electronic health records, medical images, laboratory test results, physician diagnostic reports, and patient medical records, from multiple source systems of medical institutions, such as hospital information systems, laboratory information management systems, image archiving and communication systems, and perform data cleaning to remove noise and redundant information;

[0148] S22. Build a knowledge base based on the latest credible medical information, including the latest medical data classification and grading management standards, cutting-edge papers published in important medical academic journals, and authoritative research reports;

[0149] S23. Through search enhancement technology, relevant knowledge and rules of medical data to be classified and graded are retrieved from the knowledge base, and combined with prompt engineering technology to assist the large model to achieve more accurate data classification and grading;

[0150] S24. Use the trained large model to classify and grade each piece of medical data, map them one by one to the data classification and grading rules, and add corresponding classification and grading labels.

[0151] Preferably, the step S3 specifically includes the steps of:

[0152] S31. Calculate hash values ​​for the classified and graded medical data to ensure the uniqueness and integrity of the data;

[0153] S32. Use asymmetric encryption algorithms to encrypt medical data to ensure data security during transmission and storage;

[0154] S33. Upload the encrypted medical data and its classification and grading labels to the blockchain network, record them in the distributed ledger, and verify the correctness and integrity of the data storage through the blockchain consensus mechanism to ensure that the data cannot be tampered with.

[0155] Preferably, if Figure 4 As shown, the step S4 specifically includes the following steps:

[0156] S41. The user packages the request including the request body, target data object, and the operation to be performed to generate a request R:

[0157] R←F(S, O, A),

[0158] Among them, R represents user request, S represents subject attributes (including user unique ID and permission level), O represents object attributes (including data category and level), and A represents operation attributes (including data operations such as read and write);

[0159] After signing with the private key, send the public key, certificate signature, and timestamp to the blockchain medical data management system:

[0160] B←X{PK X ,Sign(R,SK X ), T1},

[0161] Among them, B represents blockchain, X represents user, PK X Represents the public key of user X, Sign() represents the digital signature, SK X represents the private key of user X, T1,…T m Indicates a timestamp;

[0162] S42. After receiving the user request, the blockchain system uses its public key to parse the request and calls the policy management contract to automatically match the data object corresponding to the request:

[0163] B{Sign(R,SK X ), P(R), T2}→X,

[0164] Among them, P represents the policy management contract, which is responsible for automatically matching the corresponding data objects according to user requests and policy rules;

[0165] S43. Call the permission authentication contract to verify the user's access rights:

[0166] B{V(X), T3}→B,

[0167] V represents the permission verification contract, which is responsible for verifying whether the user has the permission to access a specific data object. If the verification is successful, the data retrieval contract is called to return the data corresponding to the request:

[0168] B{D(R), T4}→X,

[0169] Among them, D represents the data retrieval contract, which is responsible for retrieving and returning the data object requested by the user from the blockchain. If the verification fails, it means that the user request does not meet the policy information in the policy management contract, and a rejection message is returned:

[0170] B{Refused,T4}→X,

[0171] Refused indicates the plain text of the rejection information.

[0172] Preferably, the step S5 specifically includes the steps of:

[0173] S51. When the medical data classification and grading specifications change, reselect and label the fine-tune data set according to the new medical data classification and grading specifications, and repeat step S1;

[0174] S52. Update the knowledge base to ensure the timeliness and accuracy of retrieval information, provide the latest and most reliable knowledge support for the classification and grading management of medical data, and ensure the efficiency and practicality of retrieval enhancement technology;

[0175] S53. Update the classification and grading labels of the medical data using the updated large-scale model classification and grading architecture, and re-upload the updated data to the blockchain system;

[0176] S54. Adjust user access rights according to the updated medical data classification and grading specifications to ensure data privacy and security.

[0177] like Figure 5 As shown, another preferred embodiment of the present application further provides a medical data classification and grading security management device that combines a large model with blockchain technology, including:

[0178] The module for building a large model for medical data classification and grading is used to build a large model for medical data classification and grading. After pre-training, fine-tuning, and parameter optimization training of the large model using public medical data, the large model for medical data classification and grading is obtained.

[0179] The medical data classification and grading processing module is used for medical data classification and grading. It uses the trained medical data classification and grading model to classify and grade each piece of medical data, maps it one by one to the data classification and grading rules, and adds corresponding classification and grading labels.

[0180] The data on-chain and secure storage module is used for data on-chain and secure storage. It processes and uploads classified and graded medical data to the chain, records it in a distributed ledger, and verifies the correctness and integrity of the data storage through the blockchain consensus mechanism to ensure that the data cannot be tampered with.

[0181] Smart contract dynamic authorization and access control module, used for dynamic authorization and access control of smart contracts. After the blockchain system receives a user request, it calls the relevant smart contract to verify the user's access rights and returns the verification result;

[0182] The classification and grading specification dynamic update module is used to dynamically update the classification and grading specifications. When the classification and grading specifications of medical data change, the dynamic update process is triggered, the knowledge base is updated, and the updated data is re-uploaded to the blockchain system, user access rights are adjusted, and data privacy and security are guaranteed.

[0183] In summary, the above embodiments of the present application have the following features:

[0184] This application combines large models, model fine-tuning, and retrieval enhancement technologies to optimize the medical data classification and grading process, significantly improving efficiency and accuracy. The large model's powerful multimodal data processing capabilities reduce manpower and time costs. Model fine-tuning enables the system to quickly adapt to the characteristics of different medical data, ensuring high-precision classification and grading. Retrieval enhancement technology improves the system's adaptability and scalability by dynamically retrieving relevant knowledge and rules, providing medical institutions with an efficient and reliable data management solution and helping to unlock the value of data in the medical industry.

[0185] This application proposes a multi-institution collaborative modeling method based on federated learning, which innovatively realizes cross-institutional joint training of data that is "available but invisible", effectively breaking through the traditional medical data island problem. At the same time, the present invention combines spatial anomalies, behavioral anomalies, and amplitude anomalies to design a multi-dimensional poisoning attack detection mechanism to improve the security of the model training process, and combines the local model quality gain, the amount of data provided, the training intensity, and the number of abnormal labels to calculate the contribution of each node, dynamically adjust its contribution weight, reduce the impact of bad data, and improve the overall accuracy and robustness of the model in medical data classification and grading tasks. In particular, in the face of complex scenarios where the data quality of different medical institutions is uneven, it can still ensure excellent modeling results and has extremely high application promotion value and industry adaptability.

[0186] This application leverages blockchain's distributed ledger technology to ensure decentralized data storage, avoiding the risk of single points of failure and data loss. The combination of asymmetric encryption and hashing algorithms ensures data integrity and immutability, effectively preventing malicious tampering or forgery during transmission and storage. Furthermore, blockchain's transparency and traceability ensure that every step of data management is permanently recorded and publicly accessible, enabling traceability of management responsibilities. This provides medical institutions and regulatory authorities with a reliable audit basis and builds a transparent, trustworthy, and efficient technical framework for the secure management of medical data.

[0187] This application uses smart contract technology to set access rules based on data classification and grading labels, and deploys them on the blockchain to ensure that data of different levels is only open to users with corresponding permissions. This rule-based automated access control mechanism not only avoids operational errors and subjective biases in traditional manual management, but also ensures the strict implementation and traceability of access rules through the immutability and transparency of the blockchain. In addition, smart contracts support dynamic management of user rights and can adjust access rights in real time according to changes in user roles, data sensitivity, and business needs, effectively preventing users from exceeding their authority or abusing data. Through the synergy of smart contracts and blockchain, the present invention realizes the full process automation of data storage, access control, and rights management, providing efficient and reliable technical guarantees for the safe use of medical data, while reducing management costs and risks, and helping the medical industry build a more intelligent and secure data management system.

[0188] like Figure 6 As shown, a preferred embodiment of the present application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the medical data classification and grading security management method combining the big model and blockchain technology in the above embodiment are implemented.

[0189] like Figure 7 As shown, the preferred embodiment of the present application further provides a computer device, which can be a terminal or a liveness detection server, and its internal structure diagram can be as shown in FIG. Figure 7 As shown. The computer device includes a processor, memory, and a network interface connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with other external computer devices via a network connection. When executed by the processor, the computer program implements the steps of the above-mentioned medical data classification and grading security management method combining a large model and blockchain technology.

[0190] Those skilled in the art will understand that Figure 7 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0191] A preferred embodiment of the present application also provides a storage medium, which includes a stored program, and when the program is running, controls the device where the storage medium is located to execute the steps of the medical data classification and grading security management method combining the big model and blockchain technology in the above embodiment.

[0192] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0193] If the functions described in the method of this embodiment are implemented in the form of a software functional unit and sold or used as an independent product, they can be stored in a storage medium readable by one or more computing devices. Based on this understanding, the part of the embodiment of the present application that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes a number of instructions for enabling a computing device (which can be a personal computer, server, mobile computing device or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program code.

[0194] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the application can adopt 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.) that contain computer-usable program code. The scheme in the embodiment of the present application can be implemented in various computer languages, for example, object-oriented programming language Java and literal translation scripting language JavaScript, etc.

[0195] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0196] These computer program instructions may 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 produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0197] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0198] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present application.

[0199] Obviously, those skilled in the art may make various changes and modifications to this application without departing from the spirit and scope of this application. Thus, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application is intended to include these modifications and variations.

Claims

1. A medical data classification and grading security management method combining a large model with blockchain technology, characterized by: Including steps: S1. Construction of a large model for medical data classification and grading. After pre-training, fine-tuning, and parameter optimization training of the large model using public medical data, a large model for medical data classification and grading is obtained. S2. Classification and grading of medical data: Use the trained medical data classification and grading model to classify and grade each piece of medical data, map it one by one to the data classification and grading rules, and add corresponding classification and grading labels; S3, data on-chain and secure storage: Classified and graded medical data is processed and uploaded to the chain, recorded in a distributed ledger, and the correctness and integrity of the data storage are verified through the blockchain consensus mechanism to ensure that the data cannot be tampered with; S4. Dynamic authorization and access control of smart contracts. After receiving a user request, the blockchain system calls the relevant smart contract to verify the user's access rights and returns the verification result. S5. Dynamic update of classification and grading specifications. When the classification and grading specifications of medical data change, the dynamic update process is triggered, the knowledge base is updated, and the updated data is re-uploaded to the blockchain system, user access rights are adjusted, and data privacy and security are guaranteed.

2. The medical data classification and grading security management method combining a large model and blockchain technology according to claim 1 is characterized in that: The step S1 specifically includes the following steps: S11. Pre-train large models using a large amount of public medical data to equip them with professional general medical knowledge. S12. Based on existing policies and regulations, formulate a classification and grading framework based on the basic nature, business attributes, and potential risks of medical data; S13. Each medical institution uses data sampling techniques to collect representative small sample datasets from different categories and levels of data from its own private database, including representative diagnosis and treatment data and rare case data. The collected data are manually classified and graded through multi-expert evaluation methods to construct a high-quality fine-tuning dataset; S14. Deploy the pre-trained large model locally in each medical institution and fine-tune the pre-trained large model based on the personalized data of each medical institution through low-rank adaptation technology. Under the condition of freezing the original parameters of the pre-trained model, train it on the labeled data set and obtain the matrix of parameter changes based on the gradient descent optimization of the cross entropy loss function: Among them, y i is the true label, is the probability predicted by the model, n is the number of categories, and the fine-tuning method is applied to obtain the parameter change matrix (ΔW1,…,ΔW N ); S15, the parameter change matrix (ΔW1, ..., ΔW N ) Use homomorphic encryption and differential privacy encryption technology to mask the gradient information and send the encrypted parameter change matrix to the central aggregation server; S16. Use a method that combines spatial anomaly, behavioral anomaly, and amplitude anomaly multi-indicator detection to detect poisoning attacks; S17. Calculate the contribution of each node based on the local model quality gain, the amount of data provided and the training intensity, and the number of abnormal labels; S18. The central server performs weighted security aggregation based on the contribution of each institution. The weight change matrix of the institution with a large contribution occupies a larger weight in the training. The parameter change matrix after weighted addition is added to the parameter matrix of the pre-trained large model, and the accuracy change is recorded. When the accuracy change is less than the expected value, the model is considered to have converged and the training is terminated. Otherwise, the global model parameter change is encrypted and sent to each institution node. After updating the local model, return to step S13 until the model training converges.

3. The medical data classification and grading security management method combining a large model and blockchain technology according to claim 2 is characterized in that: The step S16 specifically includes the following steps: S161. First, the parameter change matrices uploaded by each organization are mapped to a unified vector space and the cosine similarity between them is calculated according to the following formula: The parameter changes of malicious nodes are significantly lower than the similarity of most normal nodes, forming isolated clusters. A threshold θ is set. When the average cosine similarity of node i with other nodes is less than the set threshold, it is considered an abnormal user. S162. Next, combine the parameter change matrix of each medical institution with the base large model, record the resulting performance changes, and mark the user as an abnormal user when the model performance degradation value exceeds the normal range. S163. Finally, the norm of the parameter change matrix uploaded by all nodes is counted. When a node has malicious attack behavior, a large variation of ΔW will often appear in an attempt to control the model. If the norm of a node is much larger than the mean, it will be marked as an abnormal user; warning feedback will be given to the abnormal user, requiring the node to re-check the data label or re-fine-tune, and accumulate abnormal labels for it. Its ΔW will be downgraded during aggregation. When the cumulative number exceeds the preset threshold, the organization node will be permanently eliminated.

4. The medical data classification and grading security management method combining a large model and blockchain technology according to claim 2 is characterized in that: The step S17 specifically includes: S171. Use accuracy, F1 score, ROC and other indicators to record the local gain of ΔW of each mechanism to the global model, denoted as G; S172. Count the number of labeled samples and training rounds used for fine-tuning for each medical institution, denoted as Q, to avoid "free-riding" of nodes with extremely small data sizes and very few training rounds; S173. Calculate the contribution of each medical institution based on the number of times F the node was identified as an abnormal node during training using the following formula: The weight parameters a, β, and γ can be adjusted dynamically as needed. A minimum protection threshold is set to avoid complete power loss for small medical institutions.

5. The medical data classification and grading security management method combining a large model and blockchain technology according to claim 1 is characterized in that: The step S2 specifically includes the following steps: S21. Collect structured and unstructured medical data, including electronic health records, medical images, laboratory test results, physician diagnostic reports, and patient medical records, from multiple source systems of medical institutions, such as hospital information systems, laboratory information management systems, image archiving and communication systems, and perform data cleaning to remove noise and redundant information; S22. Build a knowledge base based on the latest credible medical information, including the latest medical data classification and grading management standards, cutting-edge papers published in important medical academic journals, and authoritative research reports; S23. Through search enhancement technology, relevant knowledge and rules of medical data to be classified and graded are retrieved from the knowledge base, and combined with prompt engineering technology to assist the large model to achieve more accurate data classification and grading; S24. Use the trained large model to classify and grade each piece of medical data, map them one by one to the data classification and grading rules, and add corresponding classification and grading labels.

6. The medical data classification and grading security management method combining a large model and blockchain technology according to claim 1 is characterized in that: The step S3 specifically includes the following steps: S31. Calculate hash values ​​for the classified and graded medical data to ensure the uniqueness and integrity of the data; S32. Use asymmetric encryption algorithms to encrypt medical data to ensure data security during transmission and storage; S33. Upload the encrypted medical data and its classification and grading labels to the blockchain network, record them in the distributed ledger, and verify the correctness and integrity of the data storage through the blockchain consensus mechanism to ensure that the data cannot be tampered with.

7. The medical data classification and grading security management method combining a large model and blockchain technology according to claim 1 is characterized in that: The step S4 specifically includes the following steps: S41. The user packages the request including the request body, target data object, and the operation to be performed to generate a request R: R←F(S, O, A), Among them, R represents user request, S represents subject attributes (including user unique ID and permission level), O represents object attributes (including data category and level), and A represents operation attributes (including data operations such as read and write); After signing with the private key, send the public key, certificate signature, and timestamp to the blockchain medical data management system: B←X{PK X ,Sign(R,SK X ),T1}, Among them, B represents blockchain, X represents user, PK X Represents the public key of user X, Sign() represents the digital signature, SK X represents the private key of user X, T1,…T m Indicates a timestamp; S42. After receiving the user request, the blockchain system uses its public key to parse the request and calls the policy management contract to automatically match the data object corresponding to the request: B{Sign(R,SK X ),P(R),T2}→X, Among them, P represents the policy management contract, which is responsible for automatically matching the corresponding data objects according to user requests and policy rules; S43. Call the permission authentication contract to verify the user's access rights: B{V(X), T3}→B, V represents the permission verification contract, which is responsible for verifying whether the user has the permission to access a specific data object. If the verification is successful, the data retrieval contract is called to return the data corresponding to the request: B{D(R), T4}→X, Among them, D represents the data retrieval contract, which is responsible for retrieving and returning the data object requested by the user from the blockchain. If the verification fails, it means that the user request does not meet the policy information in the policy management contract, and a rejection message is returned: B{Refused,T4}→X, Refused indicates the plain text of the rejection information.

8. The medical data classification and grading security management method combining a large model and blockchain technology according to claim 1 is characterized in that: The step S5 specifically includes the following steps: S51. When the medical data classification and grading specifications change, reselect and label the fine-tune data set according to the new medical data classification and grading specifications, and repeat step S1; S52. Update the knowledge base to ensure the timeliness and accuracy of retrieval information, provide the latest and most reliable knowledge support for the classification and grading management of medical data, and ensure the efficiency and practicality of retrieval enhancement technology; S53. Update the classification and grading labels of the medical data using the updated large-scale model classification and grading architecture, and re-upload the updated data to the blockchain system; S54. Adjust user access rights according to the updated medical data classification and grading specifications to ensure data privacy and security.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the medical data classification and grading security management method combining a large model and blockchain technology are implemented as described in any one of claims 1 to 8.

10. A storage medium comprising a stored program, characterized in that: When the program is running, the device where the storage medium is located is controlled to execute the steps of the medical data classification and grading security management method combining a large model and blockchain technology as described in any one of claims 1 to 8.

Images