Health data encryption storage method and device, equipment and storage medium

Abstract

The invention relates to a health data encryption storage method and device, equipment and a storage medium, and the method comprises the steps: obtaining to-be-processed health data, carrying out the sensitivity analysis and grade division of the health data, and obtaining health data sub-blocks; performing encryption preprocessing on the health data sub-blocks, and performing group data protection according to a preset differential privacy algorithm to obtain encrypted data blocks; generating an initial master key through a hardware security module, and performing derived hierarchical encryption on the encrypted data block to obtain a data encryption key; performing fragmentation processing and regular distributed storage on the data encryption key to obtain a dynamic security key; and performing metadata separation storage on the encrypted data according to the dynamic security key, and performing data permission determination according to a preset role-based access control mechanism to obtain an encrypted isolation database. According to the invention, efficient security protection can be maintained under different application scenes and access permissions, and the risk of data leakage is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of encrypted storage of health data, and in particular to a method, apparatus, device and storage medium for encrypted storage of health data. Background Art

[0002] With the rapid development of information technology and the digital transformation of the health care field, the collection, storage and analysis of health data have become an indispensable part of the modern medical system. As a highly sensitive personal information, the secure storage and privacy protection of health data are increasingly attracting widespread attention from all walks of life. How to effectively protect the security and privacy of health data while ensuring data availability has become one of the key research topics. Existing health data storage methods often use a single encryption technology or static access control mechanism, which fails to fully consider the diversity and sensitivity of health data, and also ignores the changes in data security requirements in different scenarios. This simplified protection method is difficult to cope with increasingly complex network security threats, and may lead to inadequate protection of health data, thereby increasing the risk of data leakage and affecting patients' privacy rights and the credibility of medical institutions. Summary of the Invention

[0003] The main purpose of the present invention is to provide a health data encryption storage method, device, equipment and storage medium, which can ensure efficient security protection in different application scenarios and access rights, thereby reducing the risk of data leakage.

[0004] To achieve the above objectives, the present invention provides a method for encrypting and storing health data, comprising: Acquiring health data to be processed, performing sensitivity analysis and grading on the health data, and obtaining health data sub-blocks; Performing encryption preprocessing on the health data sub-blocks and performing group data protection according to a preset differential privacy algorithm to obtain encrypted data blocks; Generate an initial master key through a hardware security module, perform derived hierarchical encryption on the encrypted data block to obtain a data encryption key; The data encryption key is subjected to sharding processing and periodic distributed storage to obtain a dynamic security key; The encrypted data is stored separately from metadata based on the dynamic security key, and data permissions are identified based on a preset role-based access control mechanism to obtain an encrypted isolation database.

[0005] Furthermore, the acquiring of the health data to be processed, performing sensitivity analysis and grading on the health data to obtain health data sub-blocks, includes: Identify the data type of the acquired health data and classify the data according to the preset health data classification standards to obtain health category data; Performing a sensitivity quantitative assessment on the health category data to obtain a sensitivity value matrix; Clustering the health data according to the sensitivity value matrix to obtain initial data clusters; Calculating information redundancy and correlation of the initial data cluster to obtain the data cluster boundary; The health data is divided into blocks according to the data cluster boundaries, and identifiers and sensitivity level labels are assigned to obtain the health data sub-blocks.

[0006] Furthermore, the health data sub-block is encrypted and pre-processed, and group data protection is performed according to a preset differential privacy algorithm to obtain an encrypted data block, including: Performing a static hash verification calculation on the health data sub-block to obtain a data integrity verification identifier; Classifying the health data sub-blocks according to the data integrity verification identifier to obtain a data aggregation set; Adding random noise to the aggregated data set to obtain preliminary disturbance data; performing exponential processing on the preliminary disturbance data to obtain a multi-dimensional disturbance vector; Constructing a privacy budget allocation matrix based on the multidimensional perturbation vector, and dynamically allocating privacy budget values ​​to obtain a differential privacy protection parameter set; Performing a fusion operation on the differential privacy protection parameter set and the data aggregation set to obtain privacy-preserving intermediate data; Performing boundary decorrelation weakening on the privacy-preserving intermediate data to obtain a decorrelation data set; The de-correlated data set is encrypted block by block to obtain the encrypted data block.

[0007] Furthermore, the generating of an initial master key by the hardware security module and performing derived hierarchical encryption on the encrypted data block to obtain a data encryption key includes: Performing multiple rounds of iterations on the initial master key through a preset key derivation function to obtain a multi-level key tree structure; Performing a key hierarchy extraction on the multi-level key tree structure according to the health data to obtain a hierarchical mapping key; Performing data block hash connection on the hierarchical mapping key to obtain a data block feature vector; Performing homomorphic encryption transformation on the encrypted data block based on the data block feature vector to obtain homomorphic encrypted data; Combining and encapsulating the homomorphically encrypted data with the hierarchical mapping key to obtain an intermediate encryption key; Performing high-intensity key expansion processing on the intermediate encryption key to obtain an expanded key sequence; Performing key dispersion decomposition on the extended key sequence to obtain a main key and an auxiliary key; Time-binding the subject key according to a preset time one-way function to obtain a time-bound key; An XOR operation is performed on the time-bound key and the auxiliary key, and key encapsulation is performed to obtain the data encryption key.

[0008] Furthermore, the data encryption key is subjected to sharding processing and periodic distributed storage to obtain a dynamic security key, including: Performing polynomial threshold sharding on the data encryption key to obtain multiple key fragments; Performing unique hash identification marking on the multiple key fragments to obtain identification key fragments; Performing a timestamp embedding process on the identification key fragment to obtain a time-effective key fragment; Construct distributed storage nodes according to the predefined network topology and obtain the node communication verification matrix; Performing secondary encryption on the time-sensitive key fragment to obtain a reinforced key fragment; Allocating the reinforced key fragments to nodes according to the node communication verification matrix to obtain storage node key fragments; The storage node key fragments are repositioned according to a preset time interval, and the position information, time information and verification rules are integrated to obtain the dynamic security key.

[0009] Furthermore, the encrypted data is stored separately from metadata based on the dynamic security key, and data permissions are identified based on a preset role-based access control mechanism to obtain an encrypted isolation database, including: Separating the encrypted data from its content and attribute information to obtain private subject data and metadata; Marking the privacy subject data with a sensitivity level, and binding and associating it with the dynamic security key to obtain key binding data; Constructing a retrieval structure for the metadata to obtain indexable metadata; Randomizing the storage location of the indexable metadata to obtain a decentralized metadata database; Perform role association on the key binding data according to the role-based access control mechanism to obtain a role permission mapping table; Constructing a data access channel based on the role permission mapping table and the indexable metadata to obtain a controlled access path; The key binding data and the distributed storage metadata database are logically partitioned according to the controlled access path to obtain the encrypted isolation database.

[0010] The present invention also provides a health data encryption storage device, which is applied to any of the above health data encryption storage methods, comprising: an acquisition module, the acquisition module being used to acquire health data to be processed, perform sensitivity analysis and grade classification on the health data, and obtain health data sub-blocks; An analysis module, configured to perform encryption preprocessing on the health data sub-blocks and perform group data protection according to a preset differential privacy algorithm to obtain encrypted data blocks; an association module, the association module being configured to generate an initial master key through a hardware security module, and perform derived hierarchical encryption on the encrypted data block to obtain a data encryption key; a processing module, the processing module being configured to perform sharding processing and periodic distributed storage on the data encryption key to obtain a dynamic security key; A control module is used to separate and store the encrypted data from metadata based on the dynamic security key, and to identify data permissions based on a preset role-based access control mechanism to obtain an encrypted isolation database.

[0011] The present invention also provides a health data encryption storage device, comprising: Memory, used to store programs; The processor is used to execute the program to implement each step of the health data encryption storage method described in any one of the above.

[0012] The present invention also provides a storage medium storing computer instructions, wherein the computer instructions are used to enable a computer to execute any of the above methods.

[0013] The present invention provides a method, device, equipment, and storage medium for encrypted health data storage, which has the following beneficial effects: By performing sensitivity analysis and categorizing health data, the privacy protection requirements of different types of health data can be more accurately assessed, thereby improving data protection accuracy and providing a more reliable foundation for secure storage. By pre-encrypting health data sub-blocks and combining them with differential privacy algorithms for group data protection, refined management of health data of varying sensitivity levels is achieved, helping to ensure data availability while safeguarding privacy and avoiding functional loss caused by over-protection. Derived hierarchical encryption based on an initial master key generated by a hardware security module ensures effective security protection across different application scenarios and access rights, reducing the risk of data leakage and improving overall security. Sharding and periodically distributing data encryption keys creates a dynamic security key mechanism. Combined with metadata separation and storage, a more robust data isolation strategy is developed, effectively defending against various cyberattacks and mitigating the security risks of centralized storage. Data permission determination based on a role-based access control mechanism allows for flexible adjustment of data access policies based on the characteristics and changing needs of different user roles, making the system more adaptable to diverse application scenarios in healthcare. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 This is a flow chart of a health data encryption storage method provided by the present invention; Figure 2 This is a structural diagram of a health data encryption storage device provided by the present invention; Figure 3 This is a structural diagram of a health data encryption storage device provided by the present invention.

[0015] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0017] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0018] Reference Figure 1 As shown, the present invention provides a method for encrypting and storing health data, comprising: Step S1: Obtain the health data to be processed, perform sensitivity analysis and level classification on the health data, and obtain health data sub-blocks; Step S2: Encryption preprocessing is performed on the health data sub-block, and group data protection is performed according to a preset differential privacy algorithm to obtain an encrypted data block; Step S3: Generate an initial master key through the hardware security module, perform derived hierarchical encryption on the encrypted data block, and obtain a data encryption key; Step S4: Slice and periodically distribute the data encryption key to obtain a dynamic security key; Step S5: Separate the metadata storage of the encrypted data based on the dynamic security key, and identify the data permissions based on the preset role-based access control mechanism to obtain an encrypted isolation database.

[0019] Based on the above steps, the detailed process is as follows: Step S1: After acquiring health data, the sensitivity analysis process begins. This process determines the data's sensitivity through a multi-dimensional assessment. Assessment dimensions include data identity (personal identification information such as name and ID number), health sensitivity (such as mental health, sexual health, and genetic information), and potential risk level (the potential harm caused by a data breach). Sensitivity analysis uses machine learning algorithms to scan data items, identifying potentially sensitive information patterns and verifying compliance with regulatory requirements (such as HIPAA and GDPR). Based on the sensitivity analysis results, health data is classified according to pre-set grading standards, typically categorized as high sensitivity (such as HIV status and mental health records), medium sensitivity (such as general diagnostic results and medication records), and low sensitivity (such as basic physiological indicators and exercise data). Semantic segmentation technology is used during data segmentation to ensure that related data items are classified into the same sensitivity level, maintaining the semantic integrity of the data. After segmentation, the system partitions the data into sub-blocks, separating data of different sensitivity levels into independent sub-blocks. Each sub-block contains data elements with similar sensitivity characteristics and is assigned a unique identifier to record the associations between the original data. The sub-block segmentation process uses information entropy calculation to ensure balanced data distribution and avoid the accumulation of sensitive information. The resulting health data sub-blocks serve as the basic units for subsequent encryption processing, with encryption strategies of corresponding strengths applied to sub-blocks of different sensitivity levels.

[0020] Step S2: The health data sub-blocks enter the encryption preprocessing phase, which includes data standardization, noise control, and redundancy elimination. Data standardization converts various health indicators into a unified format for subsequent processing; noise control uses a low-pass filtering algorithm to filter out outliers; and redundancy elimination uses principal component analysis (PCA) to compress data dimensions. The preprocessed data structure is more regular, reducing the complexity of encryption computations. The differential privacy protection mechanism is then activated. Based on mathematically proven privacy protection theory, this mechanism adds carefully designed random noise to the data, making it impossible for external observers to accurately identify individual contributions. The system dynamically adjusts the differential privacy parameter ε (privacy budget) for data sub-blocks of varying sensitivity levels, with a smaller ε value providing stronger protection for highly sensitive data. The differential privacy algorithm uses a Laplace, exponential, or Gaussian mechanism to inject statistical noise into the data, ensuring that aggregate query results do not leak individual information while maintaining statistical validity. In particular, for longitudinal health data (such as time-series medical records), a trajectory-sensitive differential privacy variant is applied to address privacy risks caused by temporal correlation. After differential privacy processing, data enters the block-level encryption phase, where it is initially encrypted using the AES-256 algorithm to generate a temporary ciphertext block. The system records the differential privacy parameters and encryption parameters as metadata, which is stored in association with the encrypted data block to provide a basis for subsequent decryption and access control.

[0021] Step S3: The hardware security module (HSM) activates a true random number generator (TRNG), generating a high-entropy random number sequence using physical random sources (such as electronic noise and quantum effects). This sequence is processed using the random number processing algorithm recommended by NIST SP 800-90A to form the initial master key (Master Key). The initial master key is stored in the HSM's tamper-resistant memory, protected by physical isolation. The system implements a key derivation framework based on the ANSI X9.24-1 standard, using the HMAC-based Extract-and-Expand Key Derivation Function (HKDF) algorithm to derive multiple levels of subkeys from the master key. During the derivation process, the system constructs key derivation parameters based on factors such as the sensitivity level, usage, and access rights of the health data, ensuring the diversity and specificity of the derived keys. The derived subkeys form a hierarchical structure: the top layer is the master key that protects the Key Encryption Key (KEK); the second layer contains KEKs corresponding to each sensitivity level; and the bottom layer is the Data Encryption Key (DEK), which actually encrypts the data. The system applies a corresponding DEK to each encrypted data block for secondary encryption, creating dual encryption protection. Both KEK and DEK utilize elliptic curve cryptography algorithms (such as ECDH and ECDSA) to generate asymmetric key pairs, supporting public key infrastructure (PKI) systems. The system assigns a unique identifier to each key and establishes a key management database to record key lifecycle information. The HSM implements a regular key rotation schedule, ensuring data encryption continuity and historical access through key versioning.

[0022] Step S4: The data encryption key is processed by secret sharing algorithm fragmentation. This algorithm divides the key into n fragments based on the principle of polynomial interpolation. The recovery threshold is set to k (k < n), ensuring that the original key can only be reconstructed by collecting at least k fragments. The fragmentation process uses polynomial operations over a finite field to ensure mathematical security. The key fragmentation adopts a (k, n) threshold scheme, with typical configurations such as (3, 5) or (5, 8), balancing security and availability. After fragmentation, the system starts a heterogeneous distributed storage mechanism to disperse the key fragments to storage nodes at different physical and logical locations. The storage nodes include server clusters dispersed geographically, blockchain network nodes, trusted execution environments (TEEs), and hardware security modules (HSMs). Each storage node adopts an independent access control mechanism and authentication method to prevent the risk of single-point attacks. The system establishes a key fragmentation index table to record the fragmentation location information. The index table itself is encrypted and stored with access restrictions. After the periodic key rotation mechanism is triggered, the system automatically generates a new set of key fragments and replaces the old key fragments in batches to ensure that the key is in a dynamic state. The rotation period is dynamically adjusted according to the data sensitivity level, and the keys of highly sensitive data are updated more frequently. The key status monitoring subsystem continuously tracks the health of the key fragments. When potential threats are detected, it automatically starts the emergency key recovery and replacement process. All key operations record cryptographic evidence to form tamper-proof audit logs, providing a basis for compliance audits and security incident investigations.

[0023] Step S5: Metadata is extracted and separated from the encrypted data using a dynamic security key. Metadata includes data description information (data type, creation time, update history), association information (logical relationships and references between data), and access control attributes (owner, access rights configuration). Metadata is physically separated from the original encrypted data and stored in a separate metadata repository. The metadata itself is protected by lightweight encryption. The system establishes a metadata index structure to maintain the mapping between encrypted data and metadata, ensuring query efficiency. A multi-level permission management system is constructed based on the role-based access control (RBAC) framework, defining role templates such as doctors, nurses, researchers, and administrators, each with a specific set of permissions assigned. Permissions are granularized down to the field level for precise access control. The system integrates the attribute-based access control (ABAC) mechanism to dynamically evaluate permissions based on user attributes (department, position, qualification certification), environmental attributes (access time, location, network environment), and data attributes (sensitivity level, data type). The access policy engine executes complex rule decisions and supports conditional expressions and time constraints. The encrypted database adopts a partitioned storage architecture, storing data of different sensitivity levels in logically isolated data partitions, each with independent security policies and access control lists. The system implements a data usage audit mechanism that records all data access, including the access subject, access time, access content, and operation type. An emergency access mechanism allows for temporary escalation of access rights under specific conditions (such as medical emergencies), triggering a special approval process and fully documenting the entire process. An isolation mode ensures that even if a partial data area is compromised, it will not lead to a complete data leak, providing a multi-layered defense.

[0024] The health data encryption storage method provided by the present invention can more accurately assess the privacy protection requirements of different types of health data by performing sensitivity analysis and level classification on health data, thereby improving the accuracy of data protection and providing a more reliable foundation for secure storage. By encrypting and preprocessing the health data sub-blocks and combining them with differential privacy algorithms for group data protection, refined management of health data of different sensitivity levels is achieved, which helps to ensure data availability while protecting privacy and avoid functional loss caused by overprotection. Derived hierarchical encryption based on the initial master key generated by the hardware security module can ensure that efficient security protection can be maintained under different application scenarios and access rights, reducing the risk of data leakage and improving the overall security protection level. By sharding and regularly distributing the data encryption key to form a dynamic security key mechanism, and combining metadata separation storage technology, a more reasonable data isolation strategy is formulated, thereby effectively resisting various types of network attacks and reducing the security risks of centralized storage. Data permission is determined based on the role-based access control mechanism, and data access strategies can be flexibly adjusted according to the characteristics and changing needs of different user roles, making the system more adaptable to diverse application scenarios in the medical and health field.

[0025] In one embodiment, health data to be processed is obtained, and sensitivity analysis and classification are performed on the health data to obtain health data sub-blocks, including: Obtain health data to be processed. Health data refers to various digital information generated by individuals in medical institutions, including electronic medical records, test reports, imaging data, vital sign monitoring data, and health indicators recorded by wearable devices. Health data is obtained through medical information system interfaces or patient-authorized data import methods.

[0026] When performing sensitivity analysis and grading on acquired health data, the system identifies the data type and categorizes it according to pre-set health data classification standards. Data type identification utilizes natural language processing technology and structured data analysis methods to scan and analyze the raw health data. The pre-set health data classification standards include basic information (basic personal information such as name, gender, date of birth, and contact information), physiological indicators (vital signs such as blood pressure, blood sugar, temperature, and heart rate), disease diagnosis (diagnosed disease name, ICD code, diagnosis time, etc.), treatment records (medication records, surgical records, treatment plans, etc.), laboratory tests (various laboratory test results, imaging reports, etc.), genetics (genetic testing data, family medical history, etc.), and behavioral habits (dietary habits, exercise records, sleep status, etc.). Through a comprehensive analysis of data structural and semantic features, the system categorizes health data into the above categories, forming health category data. Health category data retains the full content of the original data but adds type label attributes.

[0027] The quantitative sensitivity assessment of health data is based on a multidimensional assessment model. This model comprehensively considers factors such as data privacy risks, clinical value, and legal protection requirements to calculate a sensitivity score for each piece of health data. The sensitivity assessment dimensions include personal identification (a score of 0-10, indicating the degree to which the data can identify an individual), health risk (a score of 0-10, indicating the impact of a data breach on an individual's health risk assessment), social impact (a score of 0-10, indicating the risk of social discrimination or psychological harm caused by a data breach), and legal protection level (a score of 0-10, indicating the protection requirements for this type of data according to relevant laws and regulations). The assessment results are generated into a sensitivity matrix with dimensions n×4, where n is the number of health data items and 4 is the number of sensitivity assessment dimensions. Each element in the sensitivity matrix represents the quantitative score of a specific data item on a specific sensitivity dimension.

[0028] The k-means clustering algorithm is used to cluster health data based on a sensitivity matrix. This algorithm uses the sensitivity matrix as input and iteratively calculates the Euclidean distance between each data point and the cluster center, grouping data with similar sensitivity characteristics into the same cluster. During the clustering process, the optimal value of the number of cluster centers, k, is determined using the silhouette coefficient method and is generally set between 3 and 5, corresponding to three to five sensitivity levels: low, medium, and high. After clustering, initial data clusters are formed, each containing health data items with similar sensitivity characteristics.

[0029] The purpose of calculating information redundancy and association for initial data clusters is to optimize data partition boundaries. Information redundancy is achieved by calculating the mutual information between data items. A high mutual information value indicates a high degree of information overlap. Association is calculated based on the semantic relevance and logical dependencies between data items, using association rule mining algorithms such as the Apriori algorithm. The calculation formula is: Information redundancy R(X, Y) = I(X; Y) / H(X, Y), where I(X; Y) is the mutual information and H(X, Y) is the joint entropy. Association A(X, Y) = sup(X∪Y) / sup(X), which represents the support ratio of data items X to Y. By setting an information redundancy threshold (typically 0.7) and an association threshold (typically 0.6), we can identify closely connected regions within data clusters and the boundaries between clusters, thereby determining the cluster boundaries.

[0030] Chunking health data according to data cluster boundaries is a key step in ultimately forming health data sub-chunks. The chunking process adheres to the following principles: data with the same sensitivity level is prioritized within the same chunk; highly correlated data items are kept in the same chunk; chunk sizes are balanced to avoid overly large or undersized chunks; and information redundancy between chunks is minimized. Each chunk is assigned a unique identifier (UID) generated using the UUID format. The sensitivity level is determined based on the average sensitivity score of the data items within the chunk, categorized as S1 (low sensitivity, average sensitivity score <5), S2 (medium sensitivity, average sensitivity score 5-7), and S3 (high sensitivity, average sensitivity score >7). The resulting health data sub-chunks contain the following attributes: chunk identifier, sensitivity level tag, original data content, data type label, and sensitivity quantification information. Health data sub-chunks serve as the basic unit for subsequent encrypted storage. Different encryption strategies and access control mechanisms will be applied to sub-chunks of varying sensitivity levels.

[0031] This embodiment performs sensitivity analysis and grading on health data. This method can accurately identify the degree of privacy risk of different types of health data, effectively prevent the problem of over-protection or under-protection of data, and improve the accuracy of data security management. The quantitative assessment of sensitivity based on the multi-dimensional assessment model realizes the objective quantification of the privacy risk of health data and avoids the protection bias caused by subjective judgment in traditional methods. The application of the K-means clustering algorithm enables data with similar sensitivity characteristics to be automatically classified, greatly improving the classification efficiency and reducing the cost of manual intervention. By calculating the information redundancy and correlation between data items, the data block boundaries are optimized, while ensuring the integrity of related data, the information overlap between data blocks is minimized, and storage efficiency is improved. The health data sub-blocks finally formed serve as the basic unit of encrypted storage, realizing the precise application of differentiated encryption strategies, which not only ensures the security of highly sensitive data, but also avoids the waste of system resources caused by over-encryption of low-sensitivity data, and balances the relationship between security and performance.

[0032] In one embodiment, the health data sub-blocks are encrypted and pre-processed, and group data protection is performed according to a preset differential privacy algorithm to obtain an encrypted data block, including: Health data sub-blocks are smaller data units divided from the original health data according to predetermined rules. These data units contain various types of patient health information, such as physiological indicators and medical records. The method begins by performing a static hash verification calculation on the health data sub-blocks. This static hash verification calculation uses a hash algorithm such as SHA-256 or MD5 to map the health data sub-blocks into a fixed-length data digest, forming a data integrity verification identifier. The data integrity verification identifier is a unique string used to verify that the data has not been tampered with, ensuring data integrity during transmission and storage.

[0033] When classifying health data sub-blocks based on data integrity verification identifiers, the system groups data sub-blocks with similar hash value characteristics into the same category, forming a data aggregation set. A data aggregation set is a collection of health data sub-blocks with similar attributes or characteristics. This classification method facilitates the precise implementation of differential privacy protection. The classification process uses the K-means clustering algorithm or decision tree classification method to divide data into multiple sets based on hash value similarity.

[0034] When adding random noise to an aggregated data set, the system generates random noise based on a Laplace or Gaussian distribution and adds this noise to each data point in the aggregated data set, generating preliminary perturbed data. Preliminary perturbed data refers to the data set after adding random noise. This perturbation makes it difficult for external attackers to restore the original data through statistical analysis. The amount of noise added is determined by the data sensitivity and the preset privacy parameter ε. A smaller value for the privacy parameter ε indicates a higher degree of privacy protection.

[0035] When performing exponential processing on the preliminary perturbation data, the system maps each data point into a higher-dimensional space. Using exponential transformations, the nonlinear characteristics of the data are enhanced, resulting in a multidimensional perturbation vector. A multidimensional perturbation vector is a high-dimensional representation of the data formed after exponential processing. This representation increases the complexity of the data and further enhances data protection. Exponential processing uses functions of the form f(x)=ex or f(x)=xn, with parameters adjusted based on the specific data characteristics.

[0036] When constructing a privacy budget allocation matrix based on a multidimensional perturbation vector, the system creates a matrix structure where each element corresponds to the privacy protection strength of a data dimension. This matrix dynamically allocates privacy budget values ​​to obtain a differential privacy parameter set. The privacy budget allocation matrix is ​​a two-dimensional array that determines the proportion of privacy resources allocated to each dimension. The differential privacy parameter set contains all the parameters required to implement differential privacy, such as sensitivity and noise ratio. Budget allocation uses an exponential decay strategy, with higher-sensitivity data receiving a larger privacy budget and lower-sensitivity data receiving a smaller budget.

[0037] When the differential privacy-preserving parameter set is combined with the aggregated data set, the system uses the parameters in the parameter set to control the proportion and method of noise addition. These parameters are then applied to the data set to generate privacy-preserving intermediate data. Privacy-preserving intermediate data is the state of data that has undergone differential privacy processing but has not yet been fully processed. At this point, the data already possesses basic privacy-preserving properties.

[0038] When performing boundary decorrelation on privacy-preserving intermediate data, the system identifies and eliminates correlations between data points, particularly those between boundary data points. This weakens the mutual directivity between data points, resulting in a decorrelated dataset. A decorrelated dataset is a collection of data with enhanced independence between data items. This process reduces the possibility of inferring other data points from one data point. Decorrelation uses K-anonymization techniques or generalization methods to merge similar data and eliminate unique features.

[0039] When performing block encryption on a de-correlated dataset, the system divides the dataset into multiple independent data blocks and encrypts them using algorithms such as AES and RSA to produce encrypted data blocks. These encrypted data blocks serve as the final encrypted storage unit and offer strong security protection against unauthorized access. Encryption uses a 256-bit key length and CBC mode to provide strong security protection.

[0040] This embodiment achieves effective verification and efficient management of data integrity by performing static hash verification calculations and classified aggregation on health data sub-blocks, significantly improving the accuracy and reliability of data processing. By adding random noise and exponential processing to form a multi-dimensional perturbation vector, a multi-level data protection mechanism is constructed to effectively prevent data leakage and privacy infringement risks. By dynamically allocating privacy budgets and constructing a privacy budget allocation matrix, precise control of the strength of differential privacy protection is achieved, ensuring differentiated protection of data with different sensitivities. Through boundary decorrelation weakening processing, the correlation between data is effectively reduced, data independence is enhanced, and privacy attacks through association analysis are prevented. Finally, through block-by-block encrypted storage, while ensuring data availability, a strict security protection system is established, providing comprehensive security protection for the storage and use of health data.

[0041] In one embodiment, an initial master key is generated by a hardware security module, and a derived hierarchical encryption is performed on the encrypted data block to obtain a data encryption key, including: A hardware security module (HSM) is a dedicated cryptographic processing unit with physical tamper-proof features. It generates a high-entropy initial master key (IMK). This IMK serves as the root key for the entire encryption system, possessing sufficient randomness and entropy, and is typically 256 bits or longer.

[0042] During multiple rounds of iterative processing of the initial master key, the system uses a preset key derivation function. A key derivation function is an algorithm that expands a single key into multiple keys, such as HKDF (HMAC-based Key Derivation Function) or PBKDF2 (Password-Based Key Derivation Function 2). During this process, the system performs multiple hash operations on the initial master key, generating subkeys at different levels with each iteration, ultimately forming a multi-level key tree structure. A multi-level key tree structure is a hierarchical key organization, where the root node is the initial master key and each child node represents a subkey derived from its parent node.

[0043] When performing hierarchical key extraction based on health data, the system analyzes the type, sensitivity, and purpose of the health data and maps it to the corresponding level in the key tree. Hierarchical mapping keys are keys in the key tree that correspond to specific types of health data and are used to protect specific categories of health data.

[0044] When the system performs a block hash join on the hierarchical mapping key, it divides the health data into fixed-size blocks, calculates a hash value for each block, and concatenates these hash values ​​with the hierarchical mapping key to form a block feature vector. A block feature vector is a mathematical representation of the unique characteristics of a data block and contains the block's hash information and the corresponding key information.

[0045] When performing homomorphic encryption transformations based on a data block's feature vector, the system processes the encrypted data block using a homomorphic encryption algorithm (such as the Paillier or BGV algorithms). Homomorphic encryption is a specialized encryption technique that allows specific algebraic operations to be performed on ciphertext without decryption. The result of this step is homomorphically encrypted data—that is, ciphertext that has been homomorphically encrypted but retains the ability to perform specific operations on it.

[0046] When combining homomorphically encrypted data with a hierarchical mapping key, the system uses a key encapsulation mechanism to bind the two together to form an intermediate encryption key. The intermediate encryption key is a temporary key during the encryption process that contains both the homomorphically encrypted data and the hierarchical mapping key.

[0047] When performing high-strength key expansion on an intermediate encryption key, the system uses a key expansion algorithm (such as AES key expansion) to expand the intermediate encryption key into a longer key sequence. An expanded key sequence is a series of subkeys generated from the intermediate encryption key using a specific algorithm to enhance encryption strength.

[0048] When performing key decomposition, the system divides the extended key sequence into two parts: the main key and the auxiliary key. The main key is the main key component in the encryption process, while the auxiliary key is a secondary key component used to assist the encryption operation.

[0049] When binding a subject key to a specific time using a preset one-way time function, the system uses a time-based one-way function (such as TOTP) to bind the subject key to a specific time. A time-based one-way function introduces a time factor into cryptographic calculations, ensuring that the key is valid only within a specific time period. A time-bound key is a key processed by a time-based one-way function and has a time-based nature.

[0050] The time-bound key is XORed with the auxiliary key and then encapsulated using a key encapsulation algorithm (such as RSA-OAEP or ECIES) to obtain the final data encryption key. The data encryption key is the final product of the entire encryption process and is used to actually encrypt health data.

[0051] This embodiment achieves multi-level security protection for health data by generating an initial master key through a hardware security module and performing derived hierarchical encryption. The establishment of a multi-level key tree structure is mapped to the type of health data, so that different categories of health data can obtain independent key protection, effectively preventing global data risks caused by the leakage of a single key. The application of homomorphic encryption transformation enables encrypted data to perform specific operations without decryption, which not only ensures data security but also meets the needs of data analysis and processing. The time binding mechanism adds timeliness control to the key, so that data access rights have the ability to be dynamically adjusted to prevent the security threat of long-term key leakage. The key dispersion decomposition and XOR combination operations enhance the complexity of the key structure and improve the resistance to cryptanalysis attacks. The design of multiple transformations and encapsulation in the overall encryption process significantly improves the anti-attack ability of health data during storage, while maintaining the scalability and compatibility of the system, which is particularly suitable for applications in the medical and health field with high security requirements.

[0052] In one embodiment, the data encryption key is sharded and periodically distributed to obtain a dynamic security key, including: Data encryption keys are processed using polynomial threshold sharding. Polynomial threshold sharding is a cryptographic method that breaks down the original key into multiple fragments. The original key can only be reconstructed when a predetermined threshold number of fragments are collected. In its implementation, Shamir's Secret Sharing algorithm is used with a (t,n) threshold scheme, where t represents the minimum number of fragments required to reconstruct the key and n represents the total number of fragments, with t ≤ n. A t-1 degree polynomial is generated, with the original key as the constant term and the remaining coefficients randomly generated. The n key fragments are obtained by calculating the polynomial's value at n different points.

[0053] The generated key fragments are uniquely hashed and assigned a unique identifier to each key fragment. A unique hash identifier is a fixed-length string calculated by combining the key fragment and its serial number using a cryptographic hash function such as SHA-256. This ensures that each key fragment has a unique and unforgeable identifier. This identification method facilitates tracking and management of key fragments in a distributed environment and increases the difficulty of forging key fragments. This step results in a uniquely identified key fragment.

[0054] The identification key fragment then undergoes timestamp embedding, embedding the current timestamp and expiration date into the identification key fragment. Timestamp embedding binds the generation time and expiration date to the identification key fragment, forming a composite structure containing time-sensitive information. This timestamp uses the UTC standard time format, is accurate to the millisecond level, and is provided by a trusted time source. The expiration date is set according to the security policy and typically ranges from 24 hours to 7 days. The time-sensitive key fragment generated in this step has a time-constrained property; key fragments that exceed the expiration date automatically expire, enhancing time-sensitive security protection.

[0055] Distributed storage nodes are constructed based on a predefined network topology. This topology refers to the connections and communication paths between storage nodes, typically employing a mesh or star structure. Inter-node communication is encrypted using the TLS protocol and employs a two-way authentication mechanism. A node communication verification matrix is ​​constructed based on factors such as inter-node communication latency, bandwidth, and physical distance. This matrix contains the communication status, security level, and verification rules for each pair of nodes, which is used for the subsequent distribution and verification of key fragments.

[0056] The aging key fragment undergoes a secondary encryption process to create a reinforced key fragment. This secondary encryption utilizes asymmetric encryption technology, using the public key of each storage node to encrypt the aging key fragment, ensuring that only the target node holding the corresponding private key can decrypt it. A high-strength encryption algorithm, such as RSA-2048 or ECC, provides an additional layer of protection for the key fragment. The reinforced key fragment generated in this step cannot be decrypted or used by unauthorized parties, even if intercepted during transmission.

[0057] The hardened key fragments are assigned to nodes based on the node communication verification matrix. The node assignment process adheres to the principle of maximum security separation, ensuring that key fragments with adjacent serial numbers are assigned to nodes that are physically farthest apart, reducing the risk of regional attacks. The assignment algorithm considers node load balancing, geographic distribution, and security levels to optimize the assignment scheme and improve overall security. After the assignment is complete, a set of storage node key fragments is obtained, each securely stored on a specific node.

[0058] The relocation of key fragments of storage nodes is triggered at preset intervals. This interval typically ranges from 4 to 24 hours and can be adjusted dynamically based on the security level. During the relocation process, key fragments are redistributed across different nodes, aging information is updated, and verification rules are recalculated. The relocation algorithm uses a pseudo-random sequence generator to generate an unpredictable relocation scheme based on the current time, state, and security seed. The location information includes the new storage node identifier and storage path; the aging information updates the validity period and timestamp of the key fragment; and the verification rules include an integrity check code and access control policy. By integrating this information, a dynamic security key is ultimately generated, achieving highly decentralized key storage and dynamic updates.

[0059] Dynamic security keys are the core security mechanism for encrypted health data storage. They are characterized by spatial and temporal dispersion, time constraints, and dynamic variability. Through the multi-layered security processing described in the above embodiments, even if an attacker gains access to some storage nodes, they cannot obtain sufficient key fragments to reconstruct the original key within the valid timeframe, effectively ensuring the secure encrypted storage of health data.

[0060] This embodiment achieves highly decentralized key management by performing polynomial threshold sharding and periodic distributed storage on data encryption keys, significantly improving the security level of health data. Using unique hash identification and timestamp embedding technology, key fragments are given unique identity and time constraints, effectively preventing the risk of illegal key copying and long-term abuse. Secondary encryption processing and a node allocation mechanism based on network topology enhance the system's ability to resist network attacks. Even if an attacker obtains some nodes, they cannot reconstruct the original key within the effective time. A location reorganization strategy based on preset time intervals enables dynamic migration of key fragments, establishing a dual-space and time protection mechanism that ensures the system remains highly secure even under sustained attacks. The entire solution eliminates the need for complex hardware infrastructure and is implemented purely in software, reducing deployment costs. The generation and management of dynamic security keys is fully automated, reducing human intervention and operational risks while improving system response speed, providing comprehensive security protection for the encrypted storage of health data.

[0061] In one embodiment, encrypted data is stored separately from metadata based on a dynamic security key, and data permissions are determined based on a preset role-based access control mechanism to obtain an encrypted isolation database, including: The encrypted storage process for health data begins with the separation of data content and attribute information. Encrypted data refers to health information that has undergone preliminary encryption and includes sensitive information such as patient physiological indicators and medical records. During this separation process, the data content (i.e., private data) is separated from the attribute information describing the data (i.e., metadata). Private data contains the actual health information, while metadata includes descriptive information such as the data's creation time, type, and source. This separation decouples data content from descriptive information, enhancing security against data leaks.

[0062] The next step is to label the privacy subject's data with a sensitivity level. Sensitivity levels are categorized as high, medium, and low, depending on the data's privacy level. High-level data includes extremely sensitive information such as genetic information and mental health records; medium-level data includes relatively sensitive information such as chronic disease history and medication records; and low-level data includes relatively less sensitive information such as routine physical examination indicators. Once labeled, the sensitivity level is associated with a dynamic security key, forming key-bound data. A dynamic security key is a regularly updated encryption key generated using a timestamp and random number generation algorithm, ensuring that even if a single key is compromised, it will not compromise the security of all data.

[0063] Constructing a metadata retrieval structure is a key step in achieving efficient data retrieval. This retrieval structure uses a B+ tree index to establish index relationships between key metadata fields, creating indexable metadata. This index relationship includes a multidimensional index structure for key fields such as patient ID, data type, and timestamp, enabling efficient locating of required information during data retrieval.

[0064] Randomizing the storage location of indexable metadata is a key step in improving data security. This randomization uses a pseudo-random number generation algorithm to randomly distribute metadata within the physical storage space, breaking the linear relationship between metadata and forming a decentralized metadata repository. This physically disperses related metadata, making it more difficult to reconstruct data in the event of unauthorized access.

[0065] Role-based access control mechanisms, which associate key-bound data with roles, are the foundation for refined permission management. Role association maps pre-defined roles (such as doctor, nurse, researcher, and patient) to key-bound data of varying levels, creating a role-permission mapping table. This table details each role's access rights to data of varying sensitivity levels, enabling fine-grained, role-based access control.

[0066] Building data access channels based on role-authority mapping tables and indexable metadata is a core step in ensuring secure data access. A data access channel is a logical access path that combines user role permissions with the data index structure to form a controlled access path. This controlled access path is dynamically generated when a user initiates a data access request, ensuring that users can only access data within their authorized scope.

[0067] The final step in achieving data isolation is logically partitioning key-bound data and distributed metadata repositories based on controlled access paths. Logical partitioning logically separates data with varying sensitivity levels and role permissions, creating an encrypted, isolated database. While these databases may share the same physical storage device, they maintain strict data isolation at the logical access level, further enhancing data security.

[0068] This embodiment achieves the decoupling of data content and attribute information by storing encrypted data separately from metadata, forming an effective security barrier in the event of a data leak and reducing the risk of sensitive health information exposure. The privacy subject data is marked with a sensitivity level and bound to a dynamic security key, establishing a multi-level data protection mechanism. Even if a single key is cracked, it will not lead to the leakage of all health data. The metadata retrieval structure is constructed using a B+ tree index method, which improves retrieval efficiency while ensuring data security and balancing the needs of security and availability. The randomization of storage locations breaks the linear relationship between metadata, increases the difficulty of data reorganization in the event of unauthorized access, and improves data security. The role-based access control mechanism realizes refined permission management of health data of different sensitivity levels, ensuring that medical staff, researchers and patients can only access information within their authority. The logically partitioned encrypted isolation database architecture achieves strict data logical isolation while maintaining efficient utilization of physical storage resources, providing differentiated security protection for health data at different levels.

[0069] Reference Figure 2 As shown, the present invention also provides a health data encryption storage device, which is applied to any of the above health data encryption storage methods, including: The acquisition module is used to obtain the health data to be processed, perform sensitivity analysis and grade classification on the health data, and obtain health data sub-blocks; An analysis module is used to perform encryption preprocessing on the health data sub-blocks and perform group data protection based on a preset differential privacy algorithm to obtain encrypted data blocks; An association module is used to generate an initial master key through a hardware security module, derive hierarchical encryption on the encrypted data block, and obtain a data encryption key; A processing module, which is used to perform sharding and periodic distributed storage on the data encryption key to obtain a dynamic security key; The control module is used to separate the metadata storage of encrypted data based on the dynamic security key, and to identify data permissions based on the preset role-based access control mechanism to obtain an encrypted isolation database.

[0070] The present invention provides a health data encryption storage device that, by performing sensitivity analysis and grading of health data, can more accurately assess the privacy protection requirements of different types of health data, thereby improving the accuracy of data protection and providing a more reliable foundation for secure storage. By encrypting and preprocessing health data sub-blocks and combining them with differential privacy algorithms for group data protection, refined management of health data of different sensitivity levels is achieved, helping to ensure data availability while protecting privacy and avoiding functional loss caused by overprotection. Derived hierarchical encryption based on the initial master key generated by the hardware security module can ensure that efficient security protection can be maintained under different application scenarios and access rights, reducing the risk of data leakage and improving the overall security protection level. By sharding and regularly distributing the data encryption key to form a dynamic security key mechanism, and combining it with metadata separation storage technology, a more reasonable data isolation strategy is formulated, thereby effectively resisting various types of network attacks and reducing the security risks of centralized storage. Data permission is determined based on a role-based access control mechanism, and data access strategies can be flexibly adjusted according to the characteristics and changing needs of different user roles, making the system more adaptable to diverse application scenarios in the medical and health field.

[0071] Reference Figure 3 As shown, the present invention also provides a health data encryption storage device, comprising: Memory, used to store programs; A processor is used to execute a program to implement each step of a health data encryption storage method as claimed in any one of claims 1 to 6.

[0072] In this embodiment, the processor and memory may be connected via a bus or other means. The memory may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as read-only memory, flash memory, hard disk, or solid-state drive. The processor may be a general-purpose processor, such as a central processing unit, a digital signal processor, an application-specific integrated circuit, or one or more integrated circuits configured to implement the embodiments of the present invention.

[0073] The present invention also provides a storage medium storing computer instructions, wherein the computer instructions are used to enable a computer to execute any of the above methods.

[0074] It should be noted that, those skilled in the art will clearly understand that, for the sake of convenience and brevity of description, the specific working processes of the above-described system and each module can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0075] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for encrypting and storing health data, characterized in that: include: Acquiring health data to be processed, performing sensitivity analysis and grading on the health data, and obtaining health data sub-blocks; Performing encryption preprocessing on the health data sub-blocks and performing group data protection according to a preset differential privacy algorithm to obtain encrypted data blocks; Generate an initial master key through a hardware security module, perform derived hierarchical encryption on the encrypted data block to obtain a data encryption key; The data encryption key is subjected to sharding processing and periodic distributed storage to obtain a dynamic security key; The encrypted data is stored separately from metadata based on the dynamic security key, and data permissions are identified based on a preset role-based access control mechanism to obtain an encrypted isolation database.

2. The method for encrypting and storing health data according to claim 1, wherein: The step of obtaining the health data to be processed, performing sensitivity analysis and level classification on the health data, and obtaining health data sub-blocks includes: Identify the data type of the acquired health data and classify the data according to the preset health data classification standards to obtain health category data; Performing a sensitivity quantitative assessment on the health category data to obtain a sensitivity value matrix; Clustering the health data according to the sensitivity value matrix to obtain initial data clusters; Calculating information redundancy and correlation of the initial data cluster to obtain the data cluster boundary; The health data is divided into blocks according to the data cluster boundaries, and identifiers and sensitivity level labels are assigned to obtain the health data sub-blocks.

3. The method for encrypting and storing health data according to claim 1, wherein: The health data sub-block is encrypted and pre-processed, and group data protection is performed according to a preset differential privacy algorithm to obtain an encrypted data block, including: Performing a static hash verification calculation on the health data sub-block to obtain a data integrity verification identifier; Classifying the health data sub-blocks according to the data integrity verification identifier to obtain a data aggregation set; Adding random noise to the aggregated data set to obtain preliminary disturbance data; performing exponential processing on the preliminary disturbance data to obtain a multi-dimensional disturbance vector; Constructing a privacy budget allocation matrix based on the multidimensional perturbation vector, and dynamically allocating privacy budget values ​​to obtain a differential privacy protection parameter set; Performing a fusion operation on the differential privacy protection parameter set and the data aggregation set to obtain privacy-preserving intermediate data; Performing boundary decorrelation weakening on the privacy-preserving intermediate data to obtain a decorrelation data set; The de-correlated data set is encrypted block by block to obtain the encrypted data block.

4. The method for encrypting and storing health data according to claim 1, wherein: The step of generating an initial master key through a hardware security module and performing derived hierarchical encryption on the encrypted data block to obtain a data encryption key includes: Performing multiple rounds of iterations on the initial master key through a preset key derivation function to obtain a multi-level key tree structure; Performing a key hierarchy extraction on the multi-level key tree structure according to the health data to obtain a hierarchical mapping key; Performing data block hash connection on the hierarchical mapping key to obtain a data block feature vector; Performing homomorphic encryption transformation on the encrypted data block based on the data block feature vector to obtain homomorphic encrypted data; Combining and encapsulating the homomorphically encrypted data with the hierarchical mapping key to obtain an intermediate encryption key; Performing high-intensity key expansion processing on the intermediate encryption key to obtain an expanded key sequence; Performing key dispersion decomposition on the extended key sequence to obtain a main key and an auxiliary key; Time-binding the subject key according to a preset time one-way function to obtain a time-bound key; An XOR operation is performed on the time-bound key and the auxiliary key, and key encapsulation is performed to obtain the data encryption key.

5. The method for encrypting and storing health data according to claim 1, wherein: The data encryption key is subjected to sharding processing and periodic distributed storage to obtain a dynamic security key, including: Performing polynomial threshold sharding on the data encryption key to obtain multiple key fragments; Performing unique hash identification marking on the multiple key fragments to obtain identification key fragments; Performing a timestamp embedding process on the identification key fragment to obtain a time-effective key fragment; Construct distributed storage nodes according to the predefined network topology and obtain the node communication verification matrix; Performing secondary encryption on the time-sensitive key fragment to obtain a reinforced key fragment; Allocating the reinforced key fragments to nodes according to the node communication verification matrix to obtain storage node key fragments; The storage node key fragments are repositioned according to a preset time interval, and the position information, time information and verification rules are integrated to obtain the dynamic security key.

6. The method for encrypting and storing health data according to claim 1, wherein: The encrypted data is stored separately from metadata based on the dynamic security key, and data authority is identified based on a preset role-based access control mechanism to obtain an encrypted isolation database, including: Separating the encrypted data from its content and attribute information to obtain private subject data and metadata; Marking the privacy subject data with a sensitivity level, and binding and associating it with the dynamic security key to obtain key binding data; Constructing a retrieval structure for the metadata to obtain indexable metadata; Randomizing the storage location of the indexable metadata to obtain a decentralized metadata database; Perform role association on the key binding data according to the role-based access control mechanism to obtain a role permission mapping table; Constructing a data access channel based on the role permission mapping table and the indexable metadata to obtain a controlled access path; The key binding data and the distributed storage metadata database are logically partitioned according to the controlled access path to obtain the encrypted isolation database.

7. A health data encryption storage device, characterized in that: The method for encrypting and storing health data as described in any one of claims 1 to 6 above comprises: an acquisition module, the acquisition module being used to acquire health data to be processed, perform sensitivity analysis and grade classification on the health data, and obtain health data sub-blocks; An analysis module, configured to perform encryption preprocessing on the health data sub-blocks and perform group data protection according to a preset differential privacy algorithm to obtain encrypted data blocks; an association module, the association module being configured to generate an initial master key through a hardware security module, and perform derived hierarchical encryption on the encrypted data block to obtain a data encryption key; a processing module, the processing module being configured to perform sharding processing and periodic distributed storage on the data encryption key to obtain a dynamic security key; A control module is used to separate and store the encrypted data from metadata based on the dynamic security key, and to identify data permissions based on a preset role-based access control mechanism to obtain an encrypted isolation database.

8. A health data encryption storage device, characterized in that: include: Memory, used to store programs; A processor is used to execute the program to implement the various steps of a health data encryption storage method as described in any one of claims 1-6.

9. A storage medium, characterized in that: Computer instructions are stored, and the computer instructions are used to make a computer execute the method according to any one of claims 1 to 6.

Images