Medical data encryption transmission method, system and equipment based on dynamic grading and national secret algorithm and medium

Abstract

The invention discloses a medical data encryption transmission method, system and device based on dynamic grading and a national secret algorithm, and a medium, mainly relates to the technical field of data encryption, and aims to solve the problems that a traditional scheme encrypts medical data uniformly and does not distinguish sensitive grades, and an existing grading encryption method depends on a fixed rule and is not high in reliability. And the data content cannot be dynamically adapted. Comprising the following steps: scanning the sensitivity level of each field in a standard medical text by using a regular rule; when the corresponding highest scanning sensitivity level is not the preset high level, deleting the field with the determined sensitivity level from the standard medical text to obtain a secondary medical text; inputting the secondary medical text into a trained natural language processing model to obtain a prediction sensitivity level of each field in the output secondary medical text; dividing the standard medical text into a plurality of data blocks with the same sensitivity level according to the sensitivity level; and according to a preset encryption algorithm and an encryption transmission channel corresponding to the sensitivity level, the data block is encrypted and sent to a preset target end.

Description

Technical Field

[0001] This application relates to the field of encrypted transmission technology, and in particular to a method, system, device and medium for encrypted transmission of medical data based on dynamic hierarchical and national cryptographic algorithms. Background Technology

[0002] With the rapid development of "Internet + Healthcare", the interconnection of medical data has become a trend, but how to ensure the security of patient medical data transmission has become an issue that needs close attention.

[0003] Current medical data encryption primarily employs traditional, uniform encryption schemes, applying the same strength of encryption protection to all medical data (such as full AES encryption). This model has significant drawbacks: on the one hand, highly sensitive data (such as genetic testing results and mental illness history) faces the risk of leakage due to insufficient encryption strength; on the other hand, low-sensitivity data (such as routine physical examination indicators) suffers from reduced transmission efficiency due to excessive encryption. Furthermore, existing hierarchical encryption methods mostly rely on static rule bases (such as keyword matching), which cannot adapt to the semantic complexity of medical text. For example, the sensitivity of the same field may change in different contexts (diabetes is low-sensitivity in blood glucose test reports but high-sensitivity in insurance claim applications), and fixed rules cannot achieve dynamic recognition. Summary of the Invention

[0004] This application provides a method, system, device, and medium for encrypted transmission of medical data based on dynamic hierarchical encryption and national cryptographic algorithms. This addresses the problems of existing solutions, which uniformly encrypt medical data without distinguishing sensitivity levels, resulting in insufficient protection for highly sensitive data or wasted performance for low-sensitive data. In addition, existing hierarchical encryption methods rely on fixed rules and cannot dynamically adapt to data content.

[0005] In a first aspect, this application provides a method for encrypted transmission of medical data based on dynamic hierarchical classification and national cryptographic algorithms, the method comprising: Configure the sensitivity levels of various data types involved in medical data, and configure the preset encryption algorithms for each sensitivity level; the sensitivity levels are divided into: preset low level, preset medium level and preset high level; Sensitivity levels and locations are labeled in historical medical texts, and the labeled historical medical texts are used as training samples; the trained natural language processing model is obtained using the training samples. Obtain the heterogeneous medical text to be sent, and process the heterogeneous medical text into a standard medical text in a preset universal format; The sensitivity level of each field in the standard medical text is scanned using regular expressions. When the highest sensitivity level of the corresponding scan is not the preset high level, the field with the determined sensitivity level is deleted from the standard medical text to obtain the secondary medical text. The secondary medical text is then input into the trained natural language processing model to obtain the predicted sensitivity level of each field in the output secondary medical text. According to the sensitivity level, the standard medical text is divided into several data blocks of the same sensitivity level; according to the preset encryption algorithm and encrypted transmission channel corresponding to the sensitivity level, the data blocks are encrypted and sent to the preset target end.

[0006] In one implementation of this application, the sensitivity levels of various data types involved in medical data are configured, and preset encryption algorithms for each sensitivity level are configured, specifically including: Through the preset interface, you can obtain the data keywords corresponding to each data type, the sensitivity level corresponding to each data type, and the preset encryption algorithm for each sensitivity level. The preset encryption algorithms include at least: symmetric encryption algorithm, block encryption algorithm, and identity encryption algorithm.

[0007] In one implementation of this application, the heterogeneous medical text to be sent is obtained, and the heterogeneous medical text is processed into standard medical text in a preset universal format, specifically including: The heterogeneous medical text is parsed into key-value pairs using an XPath parser, and these key-value pairs are then stored as standard medical text.

[0008] In one implementation of this application, standard medical text is divided into several data blocks of the same sensitivity level, specifically including: Standard medical texts are divided into several initial data blocks of the same sensitivity level. When the initial data block is less than or equal to the preset capacity, the initial data block is used as the data block; When the initial data block is larger than the preset capacity, the initial data block is split according to the preset capacity, and the split data block is used as the data block.

[0009] In one implementation of this application, the preset encryption algorithm includes at least: a symmetric encryption algorithm, a block encryption algorithm, and an identity encryption algorithm; the encrypted transmission channel is divided into a dedicated transmission channel and a normal transmission channel. Based on the preset encryption algorithm and encrypted transmission channel corresponding to the sensitivity level, the data block is encrypted and sent to the preset target end, specifically including: When the sensitivity level is preset low, a symmetric encryption algorithm is used to obtain the encrypted data block, which is then sent to the symmetric encryption acquisition interface of the preset target end through a normal transmission channel. When the sensitivity level is block encryption algorithm, the preset block size is obtained, and then the block encryption algorithm is used to obtain the encrypted data block, which is then sent to the preset target end's block encryption acquisition interface through the normal transmission channel. When the sensitivity level is identity encryption algorithm, the current sender's employee ID is obtained as the public key, and the public key and read permission are sent to the CA center. Using the public key and identity encryption algorithm, the encrypted data block is obtained and sent to the identity encryption acquisition interface of the preset target end through a dedicated line transmission channel.

[0010] In one implementation of this application, after encrypting and sending the data block to the preset target end according to the preset encryption algorithm and encrypted transmission channel corresponding to the sensitivity level, the method further includes: When the preset target end obtains the encrypted data block through the symmetric encryption acquisition interface, the data block is decrypted using the preset symmetric key. When the preset target end packet encryption acquisition interface obtains the encrypted data block, it decrypts the data block using the preset packet size and preset packet key. When the preset target end identity encryption acquisition interface obtains the encrypted data block, the preset target end sends the identity information to the CA center; When the read permission includes an identity information packet, send the private key to the preset target terminal; The decrypted data block is obtained using the private key and identity encryption algorithm.

[0011] Secondly, this application provides a medical data encryption transmission system based on dynamic hierarchical classification and national cryptographic algorithms, the system comprising: The configuration module is used to configure the sensitivity level of various data types involved in medical data, and to configure the preset encryption algorithm for each sensitivity level; the sensitivity levels are divided into: preset low level, preset medium level and preset high level. The training module is used to label the sensitivity level and location of data in historical medical texts, and uses the labeled historical medical texts as training samples; using the training samples, a trained natural language processing model is obtained. The sensitivity level module is used to acquire the heterogeneous medical text to be sent, process the heterogeneous medical text into standard medical text in a preset general format; scan the sensitivity level of each field in the standard medical text using regular expressions; when the corresponding highest sensitivity level is not the preset high level, the field with the determined sensitivity level is deleted from the standard medical text to obtain secondary medical text; input the secondary medical text into the trained natural language processing model to obtain the predicted sensitivity level of each field in the output secondary medical text; The sending module is used to divide standard medical text into several data blocks of the same sensitivity level according to the sensitivity level; and to encrypt and send the data blocks to the preset target end according to the preset encryption algorithm and encrypted transmission channel corresponding to the sensitivity level.

[0012] In one implementation of this application, the sending module includes a sending unit. This is a symmetric encryption acquisition interface used to obtain encrypted data blocks by using a symmetric encryption algorithm when the sensitivity level is preset to low, and then send them to the preset target end through a normal transmission channel. When the sensitivity level is block encryption algorithm, the preset block size is obtained, and then the block encryption algorithm is used to obtain the encrypted data block, which is then sent to the preset target end's block encryption acquisition interface through the normal transmission channel. When the sensitivity level is identity encryption algorithm, the current sender's employee ID is obtained as the public key, and the public key and read permission are sent to the CA center. Using the public key and identity encryption algorithm, the encrypted data block is obtained and sent to the identity encryption acquisition interface of the preset target end through a dedicated line transmission channel.

[0013] Thirdly, this application provides a medical data encryption transmission device based on dynamic hierarchical classification and national cryptographic algorithms, the device comprising: processor; And a memory that stores executable code, which, when executed, causes the processor to execute a medical data encryption transmission method based on dynamic hierarchical and national cryptographic algorithms, as described above.

[0014] Fourthly, this application provides a non-volatile computer storage medium storing computer instructions, which, when executed, implement a medical data encryption transmission method based on dynamic hierarchical and national cryptographic algorithms as described above.

[0015] As can be seen from the above technical solutions, this application has the following advantages: By employing a dynamic sensitivity level classification mechanism, the shortcomings of traditional "one-size-fits-all" encryption schemes are fundamentally resolved. Firstly, based on a pre-defined three-tiered sensitivity classification system (low, medium, and high), and combined with regular expression scanning and NLP model-based dual judgment, precise identification of data sensitivity at the field level is achieved. For example, fields such as patient ID numbers are automatically classified as high-sensitivity data and encrypted using identity encryption algorithms, while ordinary examination indicators are assigned to low-sensitivity levels and use lightweight encryption (symmetric encryption algorithms), ensuring a strict match between encryption strength and data value. Furthermore, different levels of encrypted data are transmitted through different channels. Through dynamic matching of encryption strength and data value (e.g., high-sensitivity data uses national cryptographic algorithms + dedicated line transmission, while low-sensitivity data uses lightweight encryption + ordinary channels), a balance between security and efficiency is achieved.

[0016] Unlike fixed-rule hierarchical methods, this application introduces a pre-trained NLP model for secondary sensitivity prediction. When there is a possibility that the regularization rule has not been fully evaluated (the corresponding highest sensitivity level of the scan is not the preset high level), the pre-trained NLP model is introduced to perform secondary sensitivity prediction. Attached Figure Description

[0017] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the description will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a medical data encryption transmission method based on dynamic hierarchical and national cryptographic algorithms provided in an embodiment of this application.

[0019] Figure 2 This is a schematic diagram of the internal structure of a medical data encryption transmission system based on dynamic hierarchical and national cryptographic algorithms, provided in an embodiment of this application.

[0020] Figure 3 This is a schematic diagram of the internal structure of a medical data encryption transmission device based on dynamic hierarchical and national cryptographic algorithms, provided in an embodiment of this application. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] Those skilled in the art should understand that the embodiments described below are merely preferred embodiments of this disclosure and do not imply that this disclosure can only be implemented through these preferred embodiments. These preferred embodiments are merely used to explain the technical principles of this disclosure and are not intended to limit the scope of protection of this disclosure. Based on the preferred embodiments provided by this disclosure, all other embodiments obtained by those skilled in the art without creative effort should still fall within the scope of protection of this disclosure.

[0023] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0024] The technical solutions proposed in the embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0025] This embodiment provides a method for encrypted transmission of medical data based on dynamic hierarchical classification and national cryptographic algorithms, such as... Figure 1 As shown in the embodiments of this application, the method mainly includes the following steps: Step 110: Configure the sensitivity level of each data type involved in medical data, and configure the preset encryption algorithm for each sensitivity level.

[0026] The sensitivity levels are divided into three categories: preset low, preset medium, and preset high.

[0027] In some embodiments, the sensitivity levels of various data types involved in medical data are configured, and preset encryption algorithms for each sensitivity level are configured, specifically including: Through the preset interface, you can obtain the data keywords corresponding to each data type, the sensitivity level corresponding to each data type, and the preset encryption algorithm for each sensitivity level. The preset encryption algorithms include at least: symmetric encryption algorithm, block encryption algorithm, and identity encryption algorithm.

[0028] It is understandable that this step, by configuring the sensitivity level and encryption algorithm of medical data in a hierarchical manner, can produce the following direct technical effects: (1) Data classification standardization: Based on data keywords, the preset sensitivity level (low / medium / high) is automatically matched to eliminate the subjective bias of manual classification. For example, patient gene data must be classified as high sensitivity level; (2) Encryption strategy differentiation: High sensitivity data is forced to use identity encryption algorithm (such as IBE), medium data uses block encryption algorithm (such as AES), and low data uses symmetric encryption algorithm (such as DES) to match computing resources with data importance; (3) Compliance guarantee: The binding relationship between sensitivity level and algorithm is solidified through preset interface to directly meet the mandatory requirements of HIPAA and other regulations for medical data encryption; (4) Operational auditability: The system automatically records the data classification results and the encryption algorithm used to provide structured logs for security audit.

[0029] Step 120: Label the data in historical medical texts with sensitivity levels and locations, and use the labeled historical medical texts as training samples; use the training samples to obtain a trained natural language processing model.

[0030] It is understandable that by labeling the location and level of sensitive data in historical medical texts and training an NLP model, the following direct technical effects can be achieved: (1) Automated recognition capability: The model can identify sensitive data in the text that is not explicitly labeled (such as ID numbers and disease names in medical records), and its accuracy depends on the coverage and labeling quality of the training samples; (2) Improved processing efficiency: The labeling of sensitive data in newly added medical texts does not need to rely entirely on manual labor, and the model can complete the initial labeling for manual review; (3) Strategy consistency: The model inherits the preset sensitivity level standards in the training samples (such as uniformly labeling "HIV positive" as high sensitivity), avoiding rule deviations caused by manual secondary classification; (4) Scalability: When new data types (such as new gene detection indicators) are added, only supplementary labeled samples and retraining of the model are needed to support recognition. All of these effects are based on the basic principles of supervised learning and do not involve unverified accuracy data.

[0031] Step 130: Obtain the heterogeneous medical text to be sent, and process the heterogeneous medical text into standard medical text in a preset general format; use regular expressions to scan the sensitivity level of each field in the standard medical text; when the corresponding highest sensitivity level is not the preset high level, delete the field with the determined sensitivity level from the standard medical text to obtain secondary medical text; input the secondary medical text into the trained natural language processing model to obtain the predicted sensitivity level of each field in the output secondary medical text.

[0032] It is understandable that this step can achieve the following direct effects through standardized processing and hierarchical filtering mechanism: (1) Format compatibility: Heterogeneous medical texts (such as PDF, XML, etc.) are unified into a structured standard format after preprocessing to ensure the enforceability of rules for subsequent sensitive field scanning; (2) Preliminary filtering efficiency: Easily identifiable sensitive fields (such as age, gender) are quickly identified through regular expression rules, reducing the computational load of subsequent NLP models; (3) Model focused processing: Secondary medical texts only retain fields that are not easily identified (such as gene data), so that model resources are concentrated on the prediction of uncertain content, reducing the misjudgment rate; (4) Hierarchical compliance: Combining the dual judgment of rules and models, it is ensured that the output meets the preset sensitivity level standard (such as high / low level classification), avoiding missed judgment or excessive desensitization caused by a single method.

[0033] Retrieve the heterogeneous medical text to be sent, and process it into a standard medical text in a preset universal format. Specifically, this includes: The heterogeneous medical text is parsed into key-value pairs using an XPath parser, and these key-value pairs are then stored as standard medical text.

[0034] It is understandable that the following direct effects are achieved through structured parsing: (1) Format unification: XPath parser can force heterogeneous medical texts such as XML / HTML (such as test reports and electronic medical records) into key-value pair structures, eliminating reading obstacles caused by differences in the original format; (2) Field traceability: Key-value pair storage explicitly retains the data source path (such as " / PatientRecord / TestResult"), which facilitates the location and auditing of sensitive fields in the future; (3) Processing compatibility: The key-value structure of standard medical text is compatible with the input requirements of most rule engines and NLP models, avoiding parsing failures caused by inconsistent formats; (4) Extension adaptability: Adding heterogeneous formats (such as JSON) only requires adjusting the XPath rules without reconstructing the processing flow.

[0035] Step 140: Divide the standard medical text into several data blocks of the same sensitivity level according to the sensitivity level; encrypt and send the data blocks to the preset target end according to the preset encryption algorithm and encryption transmission channel corresponding to the sensitivity level.

[0036] It should be noted that this step achieves the following direct technical effects through a hierarchical encryption transmission mechanism: (1) Differentiated security processing: The encryption algorithm of corresponding strength is automatically matched according to the sensitivity level of the data block to avoid resource waste or insufficient protection caused by uniform encryption; (2) Channel adaptability: High-sensitivity data blocks are transmitted through a dedicated encryption channel (such as TLS 1.3), while low-sensitivity data blocks can use the standard HTTPS channel to optimize network resource allocation; (3) Data integrity: Block encryption ensures that data of each sensitivity level is protected independently during transmission, and leakage of a single data block will not affect the security of other blocks; (4) Policy configurability: The preset rules of encryption algorithms and transmission channels support dynamic adjustment and can be flexibly changed as security standards are updated.

[0037] In some embodiments, standard medical text is divided into several data blocks of the same sensitivity level, specifically including: Standard medical texts are divided into several initial data blocks of the same sensitivity level. When the initial data block is less than or equal to the preset capacity, the initial data block is used as the data block; When the initial data block is larger than the preset capacity, the initial data block is split according to the preset capacity, and the split data block is used as the data block.

[0038] It is understandable that the following direct effects are achieved here through dynamic block partitioning strategy: (1) Storage optimization: The initial data block exceeding the limit is partitioned according to the preset capacity to avoid storage or transmission buffer overflow caused by a single data block being too large; (2) Processing consistency: Ensure that the final data blocks all meet the preset capacity limit, so that subsequent encryption, transmission and other operations have a unified processing unit; (3) Resource adaptability: Small initial data blocks do not need to be partitioned and are processed directly, reducing unnecessary computational overhead; (4) Strategy controllability: The preset capacity parameters can be flexibly adjusted according to the actual system performance (such as memory threshold, network packet size).

[0039] The preset encryption algorithms include at least: symmetric encryption algorithm, block encryption algorithm, and identity encryption algorithm. The encrypted transmission channel is divided into dedicated line transmission channel and ordinary transmission channel. Based on the preset encryption algorithm and encrypted transmission channel corresponding to the sensitivity level, the data block is encrypted and sent to the preset target end, specifically including: When the sensitivity level is preset low, a symmetric encryption algorithm is used to obtain the encrypted data block, which is then sent to the symmetric encryption acquisition interface of the preset target end through a normal transmission channel. When the sensitivity level is block encryption algorithm, the preset block size is obtained, and then the block encryption algorithm is used to obtain the encrypted data block, which is then sent to the preset target end's block encryption acquisition interface through the normal transmission channel. When the sensitivity level is identity encryption algorithm, the current sender's employee ID is obtained as the public key, and the public key and read permission are sent to the CA center. Using the public key and identity encryption algorithm, the encrypted data block is obtained and sent to the identity encryption acquisition interface of the preset target end through a dedicated line transmission channel.

[0040] It is understandable that the following direct effects are achieved through the multi-level encryption transmission mechanism: (1) Algorithm adaptability: The encryption method is automatically matched according to the sensitivity level (symmetric algorithm to handle low sensitive data, group algorithm to adapt to structured data, and identity encryption to ensure access control of high sensitive data), avoiding algorithm abuse or insufficient protection; (2) Channel selectivity: Dedicated line channel ensures physical isolation of identity encrypted data, and ordinary channel carries regular encrypted data to save costs; (3) Permission accuracy: Identity encryption is combined with the employee number public key and CA center authorization to ensure that only the target end with read permission can decrypt the data; (4) Interface isolation: The three types of encrypted data are transmitted to independent interfaces to prevent key confusion or permission overstepping during the decryption process.

[0041] After encrypting and sending the data block to the preset target end according to the preset encryption algorithm and encrypted transmission channel corresponding to the sensitivity level, the method also includes: When the preset target end obtains the encrypted data block through the symmetric encryption acquisition interface, the data block is decrypted using the preset symmetric key. When the preset target end packet encryption acquisition interface obtains the encrypted data block, it decrypts the data block using the preset packet size and preset packet key. When the preset target end identity encryption acquisition interface obtains the encrypted data block, the preset target end sends the identity information to the CA center; When the read permission includes an identity information packet, send the private key to the preset target terminal; The decrypted data block is obtained using the private key and identity encryption algorithm.

[0042] As described above, this embodiment fundamentally solves the drawbacks of the "one-size-fits-all" approach in traditional encryption schemes through a dynamic sensitivity level classification mechanism. First, based on a preset three-level sensitivity classification system (low, medium, and high), and combined with regular rule scanning and NLP model dual judgment, accurate identification of data sensitivity at the field level is achieved. For example, fields such as patient ID numbers are automatically classified as high-sensitivity data and encrypted using an identity encryption algorithm, while ordinary examination indicators are assigned to low-sensitivity levels and use lightweight encryption (symmetric encryption algorithm), ensuring a strict match between encryption strength and data value. Furthermore, different levels of encrypted data are transmitted through different channels. Through dynamic matching of encryption strength and data value (e.g., high-sensitivity data uses national cryptographic algorithms + dedicated line transmission, while low-sensitivity data uses lightweight encryption + ordinary channels), a balance between security and efficiency is achieved. Unlike fixed-rule classification methods, this application introduces a pre-trained NLP model for secondary sensitivity prediction. When there is a possibility that the regular rule has not fully judged (the corresponding highest sensitivity level of the scan is not the preset high level), the pre-trained NLP model is introduced for secondary sensitivity prediction.

[0043] In addition, this application Figure 2 This application provides an embodiment of a medical data encryption transmission system based on dynamic hierarchical classification and national cryptographic algorithms. For example... Figure 2 As shown in the embodiments of this application, the system mainly includes: The configuration module 210 is used to configure the sensitivity level of various data types involved in medical data and to configure the preset encryption algorithm for each sensitivity level; among which, the sensitivity levels are divided into: preset low level, preset medium level and preset high level.

[0044] It is understandable that the configuration module 210 can produce the following direct technical effects by configuring the sensitivity level and encryption algorithm of medical data in a hierarchical manner: (1) Data classification standardization: Based on data keywords, the preset sensitivity level (low / medium / high) is automatically matched to eliminate the subjective bias of manual classification. For example, patient gene data must be classified as high sensitivity level; (2) Encryption strategy differentiation: High sensitivity data is forced to use identity encryption algorithm (such as IBE), medium data uses block encryption algorithm (such as AES), and low data uses symmetric encryption algorithm (such as DES) to match computing resources with data importance; (3) Compliance guarantee: The binding relationship between sensitivity level and algorithm is solidified through preset interface to directly meet the mandatory requirements of HIPAA and other regulations for medical data encryption; (4) Operational auditability: The system automatically records the data classification results and the encryption algorithm used to provide structured logs for security audit.

[0045] Training module 220 is used to label the sensitivity level and location of data in historical medical texts, and to use the labeled historical medical texts as training samples; using the training samples, a trained natural language processing model is obtained.

[0046] It is understandable that by labeling the location and level of sensitive data in historical medical texts and training the NLP model, the training module 220 can produce the following direct technical effects: (1) Automated recognition capability: The model can recognize sensitive data in the text that is not explicitly labeled (such as ID numbers and disease names in medical records), and its accuracy depends on the coverage and labeling quality of the training samples; (2) Improved processing efficiency: The labeling of sensitive data in newly added medical texts does not need to rely entirely on manual labor, and the model can complete the initial labeling for manual review; (3) Strategy consistency: The model inherits the preset sensitivity level standards in the training samples (such as uniformly labeling "HIV positive" as high sensitivity), avoiding rule deviations in manual secondary classification; (4) Scalability: When new data types (such as new gene detection indicators) are added, only supplementary labeled samples and retraining of the model are needed to support recognition. All of these effects are based on the basic principles of supervised learning and do not involve unverified accuracy data.

[0047] The grading module 230 is used to acquire the heterogeneous medical text to be sent, process the heterogeneous medical text into standard medical text in a preset general format, scan the sensitivity level of each field in the standard medical text using regular expressions, and delete the field with the determined sensitivity level from the standard medical text when the corresponding highest sensitivity level is not the preset high level, thus obtaining secondary medical text. The secondary medical text is then input into a trained natural language processing model to obtain the predicted sensitivity level of each field in the output secondary medical text.

[0048] It is understandable that the grading module 230 can achieve the following direct effects through standardized processing and graded filtering mechanism: (1) Format compatibility: Heterogeneous medical texts (such as PDF, XML, etc.) are unified into a structured standard format after preprocessing, ensuring the enforceability of rules for subsequent sensitive field scanning; (2) Preliminary filtering efficiency: Easily identifiable sensitive fields (such as age, gender) are quickly identified through regular rules, reducing the computational load of subsequent NLP models; (3) Model focused processing: Secondary medical texts only retain fields that are not easily identified (such as gene data), so that model resources are concentrated on the prediction of uncertain content, reducing the misjudgment rate; (4) Graded compliance: Combining the dual judgment of rules and models, it ensures that the output conforms to the preset sensitivity level standard (such as high / low level classification), avoiding missed judgments or excessive desensitization caused by a single method.

[0049] The sending module 240 is used to divide the standard medical text into several data blocks of the same sensitivity level according to the sensitivity level; and to encrypt and send the data blocks to the preset target end according to the preset encryption algorithm and encrypted transmission channel corresponding to the sensitivity level.

[0050] It should be noted that the sending module 240 achieves the following direct technical effects through the hierarchical encryption transmission mechanism: (1) Differentiated security processing: The encryption algorithm with corresponding strength is automatically matched according to the sensitivity level of the data block to avoid resource waste or insufficient protection caused by uniform encryption; (2) Channel adaptability: High-sensitivity data blocks are transmitted through a dedicated encryption channel (such as TLS 1.3), while low-sensitivity data blocks can use the standard HTTPS channel to optimize network resource allocation; (3) Data integrity: Block encryption ensures that data of each sensitivity level is protected independently during transmission, and leakage of a single data block will not affect the security of other blocks; (4) Policy configurability: The preset rules of encryption algorithm and transmission channel support dynamic adjustment and can be flexibly changed as security standards are updated.

[0051] The transmitting module 240 includes a transmitting unit. This is a symmetric encryption acquisition interface used to obtain encrypted data blocks by using a symmetric encryption algorithm when the sensitivity level is preset to low, and then send them to the preset target end through a normal transmission channel. When the sensitivity level is block encryption algorithm, the preset block size is obtained, and then the block encryption algorithm is used to obtain the encrypted data block, which is then sent to the preset target end's block encryption acquisition interface through the normal transmission channel.

[0052] When the sensitivity level is identity encryption algorithm, the current sender's employee ID is obtained as the public key, and the public key and read permission are sent to the CA center. Using the public key and identity encryption algorithm, the encrypted data block is obtained and sent to the identity encryption acquisition interface of the preset target end through a dedicated line transmission channel.

[0053] As described above, this embodiment fundamentally solves the drawbacks of the "one-size-fits-all" approach in traditional encryption schemes through a dynamic sensitivity level classification mechanism. First, based on a preset three-level sensitivity classification system (low, medium, and high), and combined with regular rule scanning and NLP model dual judgment, accurate identification of data sensitivity at the field level is achieved. For example, fields such as patient ID numbers are automatically classified as high-sensitivity data and encrypted using an identity encryption algorithm, while ordinary examination indicators are assigned to low-sensitivity levels and use lightweight encryption (symmetric encryption algorithm), ensuring a strict match between encryption strength and data value. Furthermore, different levels of encrypted data are transmitted through different channels. Through dynamic matching of encryption strength and data value (e.g., high-sensitivity data uses national cryptographic algorithms + dedicated line transmission, while low-sensitivity data uses lightweight encryption + ordinary channels), a balance between security and efficiency is achieved. Unlike fixed-rule classification methods, this application introduces a pre-trained NLP model for secondary sensitivity prediction. When there is a possibility that the regular rule has not fully judged (the corresponding highest sensitivity level of the scan is not the preset high level), the pre-trained NLP model is introduced for secondary sensitivity prediction.

[0054] The above are method embodiments of this application. Based on the same inventive concept, this application also provides a medical data encryption transmission device based on dynamic hierarchical classification and national cryptographic algorithms. Figure 3 As shown, the device includes: a processor; and a memory storing executable code thereon, which, when executed, causes the processor to execute a medical data encryption transmission method based on dynamic hierarchical and national cryptographic algorithms as described in the above embodiment.

[0055] Specifically, the server configures the sensitivity levels of various data types involved in the medical data and sets preset encryption algorithms for each sensitivity level. The sensitivity levels are divided into: preset low, preset medium, and preset high. Sensitivity levels and locations are labeled in historical medical texts, and these labeled historical medical texts are used as training samples. A trained natural language processing model is obtained using these training samples. Heterogeneous medical texts to be sent are acquired and processed into standard medical texts in a preset universal format. Regular expressions are used to scan the sensitivity levels of each field in the standard medical texts. When the highest sensitivity level detected is not the preset high, the field with the determined sensitivity level is deleted from the standard medical text, resulting in secondary medical text. The secondary medical text is input into the trained natural language processing model to obtain the predicted sensitivity levels of each field in the output secondary medical text. According to the sensitivity levels, the standard medical texts are divided into several data blocks with the same sensitivity level. Based on the preset encryption algorithm and encrypted transmission channel corresponding to the sensitivity level, the data blocks are encrypted and sent to the preset target end.

[0056] As described above, this embodiment fundamentally solves the drawbacks of the "one-size-fits-all" approach in traditional encryption schemes through a dynamic sensitivity level classification mechanism. First, based on a preset three-level sensitivity classification system (low, medium, and high), and combined with regular rule scanning and NLP model dual judgment, accurate identification of data sensitivity at the field level is achieved. For example, fields such as patient ID numbers are automatically classified as high-sensitivity data and encrypted using an identity encryption algorithm, while ordinary examination indicators are assigned to low-sensitivity levels and use lightweight encryption (symmetric encryption algorithm), ensuring a strict match between encryption strength and data value. Furthermore, different levels of encrypted data are transmitted through different channels. Through dynamic matching of encryption strength and data value (e.g., high-sensitivity data uses national cryptographic algorithms + dedicated line transmission, while low-sensitivity data uses lightweight encryption + ordinary channels), a balance between security and efficiency is achieved. Unlike fixed-rule classification methods, this application introduces a pre-trained NLP model for secondary sensitivity prediction. When there is a possibility that the regular rule has not fully judged (the corresponding highest sensitivity level of the scan is not the preset high level), the pre-trained NLP model is introduced for secondary sensitivity prediction.

[0057] In addition, this application embodiment also provides a non-volatile computer storage medium storing executable instructions, which, when executed, implement the above-described method for encrypted transmission of medical data based on dynamic hierarchical and national cryptographic algorithms.

[0058] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for encrypted transmission of medical data based on dynamic hierarchical classification and national cryptographic algorithms, characterized in that, The method includes: Configure the sensitivity levels of various data types involved in medical data, and configure the preset encryption algorithms for each sensitivity level; the sensitivity levels are divided into: preset low level, preset medium level and preset high level; Sensitivity levels and locations are labeled in historical medical texts, and the labeled historical medical texts are used as training samples; the trained natural language processing model is obtained using the training samples. The process involves: acquiring the heterogeneous medical text to be sent; processing the heterogeneous medical text into a standard medical text in a preset general format; scanning the sensitivity level of each field in the standard medical text using regular expressions; deleting the field with the determined sensitivity level from the standard medical text when the corresponding highest sensitivity level is not the preset high level, thus obtaining a secondary medical text; and inputting the secondary medical text into a trained natural language processing model to obtain the predicted sensitivity level of each field in the output secondary medical text. According to the sensitivity level, the standard medical text is divided into several data blocks of the same sensitivity level; according to the preset encryption algorithm and encrypted transmission channel corresponding to the sensitivity level, the data blocks are encrypted and sent to the preset target end.

2. The method for encrypted transmission of medical data based on dynamic hierarchical structure and national cryptographic algorithm according to claim 1, characterized in that, Configure the sensitivity levels of various data types involved in medical data, and configure the preset encryption algorithms for each sensitivity level, specifically including: Through the preset interface, you can obtain the data keywords corresponding to each data type, the sensitivity level corresponding to each data type, and the preset encryption algorithm for each sensitivity level. The preset encryption algorithms include at least: symmetric encryption algorithm, block encryption algorithm, and identity encryption algorithm.

3. The method for encrypted transmission of medical data based on dynamic hierarchical structure and national cryptographic algorithm according to claim 1, characterized in that, Retrieve the heterogeneous medical text to be sent, and process it into a standard medical text in a preset universal format. Specifically, this includes: The heterogeneous medical text is parsed into key-value pairs using an XPath parser, and these key-value pairs are then stored as standard medical text.

4. The method for encrypted transmission of medical data based on dynamic hierarchical structure and national cryptographic algorithm according to claim 1, characterized in that, Based on sensitivity levels, standard medical texts are divided into several data blocks of the same sensitivity level, specifically including: Standard medical texts are divided into several initial data blocks of the same sensitivity level. When the initial data block is less than or equal to the preset capacity, the initial data block is used as the data block; When the initial data block is larger than the preset capacity, the initial data block is split according to the preset capacity, and the split data block is used as the data block.

5. The medical data encryption transmission method based on dynamic hierarchical and national cryptographic algorithms according to claim 1, characterized in that, The preset encryption algorithms include at least: symmetric encryption algorithm, block encryption algorithm, and identity encryption algorithm. The encrypted transmission channel is divided into dedicated line transmission channel and ordinary transmission channel. Based on the preset encryption algorithm and encrypted transmission channel corresponding to the sensitivity level, the data block is encrypted and sent to the preset target end, specifically including: When the sensitivity level is preset low, a symmetric encryption algorithm is used to obtain the encrypted data block, which is then sent to the symmetric encryption acquisition interface of the preset target end through a normal transmission channel. When the sensitivity level is block encryption algorithm, the preset block size is obtained, and then the block encryption algorithm is used to obtain the encrypted data block, which is then sent to the preset target end's block encryption acquisition interface through the normal transmission channel. When the sensitivity level is identity encryption algorithm, the current sender's employee ID is obtained as the public key, and the public key and read permission are sent to the CA center. Using the public key and identity encryption algorithm, the encrypted data block is obtained and sent to the identity encryption acquisition interface of the preset target end through a dedicated line transmission channel.

6. The method for encrypted transmission of medical data based on dynamic hierarchical structure and national cryptographic algorithm according to claim 5, characterized in that, After encrypting and sending the data block to the preset target end according to the preset encryption algorithm and encrypted transmission channel corresponding to the sensitivity level, the method further includes: When the preset target end obtains the encrypted data block through the symmetric encryption acquisition interface, the data block is decrypted using the preset symmetric key. When the preset target end packet encryption acquisition interface obtains the encrypted data block, it decrypts the data block using the preset packet size and preset packet key. When the preset target end identity encryption acquisition interface obtains the encrypted data block, the preset target end sends the identity information to the CA center; When the read permission includes an identity information packet, send the private key to the preset target terminal; The decrypted data block is obtained using the private key and identity encryption algorithm.

7. A medical data encryption transmission system based on dynamic hierarchical classification and national cryptographic algorithms, characterized in that, The system includes: The configuration module is used to configure the sensitivity level of various data types involved in medical data, and to configure the preset encryption algorithm for each sensitivity level; the sensitivity levels are divided into: preset low level, preset medium level and preset high level. The training module is used to label the sensitivity level and location of data in historical medical texts, and uses the labeled historical medical texts as training samples; using the training samples, a trained natural language processing model is obtained. The sensitivity level module is used to acquire the heterogeneous medical text to be sent, process the heterogeneous medical text into standard medical text in a preset general format; scan the sensitivity level of each field in the standard medical text using regular expressions; when the corresponding highest sensitivity level is not the preset high level, the field with the determined sensitivity level is deleted from the standard medical text to obtain secondary medical text; input the secondary medical text into the trained natural language processing model to obtain the predicted sensitivity level of each field in the output secondary medical text; The sending module is used to divide standard medical text into several data blocks of the same sensitivity level according to the sensitivity level; and to encrypt and send the data blocks to the preset target end according to the preset encryption algorithm and encrypted transmission channel corresponding to the sensitivity level.

8. The medical data encryption transmission system based on dynamic hierarchical and national cryptographic algorithms according to claim 7, characterized in that, The sending module includes a sending unit. This is a symmetric encryption acquisition interface used to obtain encrypted data blocks by using a symmetric encryption algorithm when the sensitivity level is preset to low, and then send them to the preset target end through a normal transmission channel. When the sensitivity level is block encryption algorithm, the preset block size is obtained, and then the block encryption algorithm is used to obtain the encrypted data block, which is then sent to the preset target end's block encryption acquisition interface through the normal transmission channel. When the sensitivity level is identity encryption algorithm, the current sender's employee ID is obtained as the public key, and the public key and read permission are sent to the CA center. Using the public key and identity encryption algorithm, the encrypted data block is obtained and sent to the identity encryption acquisition interface of the preset target end through a dedicated line transmission channel.

9. A medical data encryption transmission device based on dynamic hierarchical classification and national cryptographic algorithms, characterized in that, The device includes: processor; And a memory storing executable code, which, when executed, causes the processor to perform a medical data encryption transmission method based on dynamic hierarchical and national cryptographic algorithms as described in any one of claims 1-6.

10. A non-volatile computer storage medium, characterized in that, It stores computer instructions, which, when executed, implement a medical data encryption transmission method based on dynamic hierarchical and national cryptographic algorithms as described in any one of claims 1-6.

Images