Self-service registration method and system for entry-exit health certificate

Abstract

The invention relates to the technical field of data processing, and discloses a self-service registration method and system for entry and exit health certificates. The method comprises the following steps: collecting multi-modal biological characteristics to generate a digital identity packet; dynamically generating a form based on the identity packet and the destination health requirement, and obtaining health declaration information to form a health data packet; querying a physical examination knowledge graph by using the health data packet, and matching items to generate a physical examination path graph; performing quality control on the distributed medical equipment data, and processing an abnormal value by applying a self-correction model; the physical examination report is subjected to multi-level auditing to generate a health certificate decision packet; and multi-channel distribution and verification of the electronic health certificate are realized. According to the invention, full-process digitization and intelligentization are realized, the handling efficiency is improved, manual intervention is reduced, the data accuracy is ensured, and cross-system verification of the health certificate is supported.

Description

Technical Field

[0001] This application relates to the technical field of data processing, and particularly to a self-service registration method and system for exit-entry health certificates. Background Art

[0002] As a key link in the infectious disease surveillance system for entry-exit personnel, exit-entry health certificates have become essential documents for people engaged in cross-border work, studying abroad, business exchanges, etc. In the current context of the continuous advancement of globalization and the increasingly frequent international personnel flow, the demand for convenient processing of health certificates is becoming increasingly prominent. The traditional exit-entry physical examination mode mostly relies on manual operations and has many drawbacks. In the information registration link, applicants need to consult or fill in forms on-site, register and input information on-site, etc., which not only consumes time and energy but also easily leads to information errors or omissions. During peak periods, the physical examination site is often overcrowded, and applicants need to queue for a long time, resulting in a very poor experience.

[0003] The deficiencies in the existing technologies are mainly reflected in the following aspects: The existing self-service registration devices have single functions, mostly only supporting the entry of basic information, unable to achieve real-time docking with medical examination data, and unable to meet the differentiated health requirements of different destinations; Secondly, there is a lack of intelligent physical examination path planning, resulting in applicants having to queue multiple times during the physical examination process, and the process is cumbersome; Thirdly, the problem of data islands is serious, and the information systems among physical examination centers, exit-entry administration departments, and health and health departments are independent of each other, and the data sharing mechanism is imperfect; Fourthly, the quality control of physical examination data relies on manual experience, lacking automated abnormal detection and correction means, which affects the accuracy of health certificates; The issuance of health certificates still mainly relies on physical certificates, and the popularization and application of electronic health certificates are insufficient, and the verification channels are single. Summary of the Invention

[0004] This application provides a self-service registration method and system for exit-entry health certificates, which is used to realize the whole process of digitization and intelligence, improve the processing efficiency, reduce manual intervention, ensure data accuracy, and support cross-system verification of health certificates.

[0005] In a first aspect, the present application provides a self-service registration method for an entry-exit health certificate. The self-service registration method for an entry-exit health certificate includes: collecting and processing multi-modal biometric information of a user to generate a digital identity package; performing dynamic form generation processing based on the user's basic information and destination area health requirement data in the digital identity package to obtain health declaration information, and integrating the health declaration information with the digital identity package to generate a health data package; performing query processing on a physical examination knowledge graph database based on the health data package, matching a set of physical examination items and generating a physical examination path map, and integrating the physical examination path map with the health data package to generate a physical examination execution package; performing real-time quality control processing on inspection data collected by distributed medical devices according to the physical examination execution package, identifying outliers in the inspection data using a multi-source heterogeneous data self-correction model, generating a structured physical examination result data set, and generating a physical examination report package; inputting the physical examination report package into a multi-level review mechanism for compliance analysis processing, generating a review result and an electronic health certificate, and generating a health certificate decision package; and performing certificate issuance and cross-system data synchronization processing based on the health certificate decision package to achieve multi-channel distribution and verification of the electronic health certificate.

[0006] In a second aspect, the present application provides a self-service registration system for an entry-exit health certificate. The self-service registration system for an entry-exit health certificate includes: A collection module, configured to collect and process multi-modal biometric information of a user to generate a digital identity package; An integration module, configured to perform dynamic form generation processing based on the user's basic information and destination area health requirement data in the digital identity package to obtain health declaration information, and integrate the health declaration information with the digital identity package to generate a health data package; A query module, configured to perform query processing on a physical examination knowledge graph database based on the health data package, match a set of physical examination items and generate a physical examination path map, and integrate the physical examination path map with the health data package to generate a physical examination execution package; A control module, configured to perform real-time quality control processing on inspection data collected by distributed medical devices according to the physical examination execution package, identify outliers in the inspection data using a multi-source heterogeneous data self-correction model, generate a structured physical examination result data set, and generate a physical examination report package; An analysis module, configured to input the physical examination report package into a multi-level review mechanism for compliance analysis processing, generate a review result and an electronic health certificate, and generate a health certificate decision package; A synchronization module, configured to perform certificate issuance and cross-system data synchronization processing based on the health certificate decision package to achieve multi-channel distribution and verification of the electronic health certificate.

[0007] In a third aspect, a self-service registration device for an entry-exit health certificate is provided, including: a memory and at least one processor, wherein instructions are stored in the memory; the at least one processor invokes the instructions in the memory to enable the self-service registration device for an entry-exit health certificate to execute the above-mentioned self-service registration method for an entry-exit health certificate.

[0008] In a fourth aspect, a computer-readable storage medium is provided, in which instructions are stored. When it runs on a computer, it enables the computer to execute the above-mentioned self-service registration method for an entry-exit health certificate.

[0009] In the technical solution provided by this application, through the collection and processing of the user's multi-modal biometric information, combined with the deep neural network model for feature extraction, a unique identifier is generated, achieving high accuracy and security in identity verification and effectively preventing identity fraud; based on the combination of the digital identity package and the destination area health requirement data, a dynamic form is generated through the decision tree algorithm, enabling the form content to accurately match the differentiated health requirements of different destinations, greatly improving the user experience and filling efficiency; by using the physical examination knowledge graph database combined with the knowledge reasoning engine for intelligent matching of physical examination items, and identifying health risk factors through machine learning algorithms, and generating a physical examination path map in combination with the shortest path algorithm, it effectively solves the problem of time waste caused by repeated examinations and unreasonable paths in traditional physical examinations; especially in the aspect of medical data quality control, an innovative multi-source heterogeneous data self-correction model is adopted, which can identify outliers in the examination data in real time. If there are abnormalities, it prompts further reexamination. Through the Bayesian inference method, targeted processing of different types of outliers is carried out, significantly improving the accuracy and reliability of the physical examination data; in the review link, a multi-level review mechanism combined with the medical knowledge graph and the expert rule base is adopted to realize the intelligence and high efficiency of the review process and reduce the subjectivity of manual review; through the PKI technology, digital signatures of electronic health certificates are carried out and distributed through multiple channels, associated with the blockchain verification service, which not only ensures the authenticity and immutability of the certificates, but also provides a convenient online verification channel to meet the mutual recognition requirements of cross-border health certificates; overall, the contribution of applying artificial intelligence algorithms and models in the specific functional field of this invention is mainly reflected in: the deep neural network improves the identity verification accuracy rate, the knowledge graph combined with the reasoning engine realizes intelligent matching of physical examination items, the multi-source heterogeneous data self-correction model solves the problem of medical data quality control, and the multi-level machine learning review mechanism replaces the traditional manual review. The comprehensive application of these AI technologies enables the entire entry-exit health certificate processing process to achieve digital transformation and intelligent upgrade, completely solving the problems of low efficiency, many human errors, data islands and difficult cross-system verification in the traditional method. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0011] Figure 1 It is a schematic diagram of an embodiment of the self-service registration method for the entry-exit health certificate in the embodiments of the present application; Figure 2 It is a schematic diagram of an embodiment of the self-service registration system for the entry-exit health certificate in the embodiments of the present application; Figure 3 It is a structural schematic block diagram of the self-service registration device for the entry-exit health certificate in the embodiments of the present invention. Detailed implementation manners

[0012] The embodiments of the present application provide a self-service registration method and system for the entry-exit health certificate. The terms first, second, third, fourth, etc. (if any) in the specification, claims, and the above drawings of the present application are used to distinguish similar objects and do not necessarily describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances. In addition, the term "including" or "having" and any deformation thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device including a series of steps or units does not necessarily limit to those clearly listed steps or units, but may include other steps or units not clearly listed or inherent to these processes, methods, products, or devices.

[0013] For ease of understanding, the following describes the specific process of the embodiments of the present application. Please refer to Figure 1 , an embodiment of the self-service registration method for the entry-exit health certificate in the embodiments of the present application includes: Step S101, collect and process the multi-modal biometric information of the user to generate a digital identity package; Step S102, perform dynamic form generation processing according to the user's basic information in the digital identity package and the health requirement data of the destination area, obtain the health declaration information, and integrate the health declaration information with the digital identity package to generate a health data package; Step S103, perform query processing on the physical examination knowledge graph database based on the health data package, match the physical examination item set and generate a physical examination path map, and fuse the physical examination path map with the health data package to generate a physical examination execution package; Step S104: Perform real-time quality control processing on the inspection data collected by distributed medical devices according to the physical examination execution package, identify outliers in the inspection data using a multi-source heterogeneous data self-correction model, generate a structured physical examination result dataset, and generate a physical examination report package; Step S105: Input the physical examination report package into a multi-level review mechanism for compliance analysis and processing, generate a review result and an electronic health certificate, and generate a health certificate decision package; Step S106: Based on the health certificate decision package, perform certificate issuance and cross-system data synchronization processing to achieve multi-channel distribution and verification of the electronic health certificate.

[0014] It can be understood that the execution subject of this application can be a self-service registration system for entry-exit health certificates, or a terminal or a server. Specifically, it is not limited here. This application embodiment is described by taking the server as the execution subject as an example.

[0015] Specifically, the face information, iris features, and fingerprint information of the user are obtained through a biometric collection device to form multi-modal biometric data. These data are then input into a deep neural network model for feature extraction. This model has multiple convolutional layers and fully connected layers and can extract the key points and texture information of various biometric features. Through a feature fusion algorithm, the feature vectors of different modalities are integrated to generate a unique identifier. When the user performs identity verification, the system compares this unique identifier with the pre-stored biometric template to generate an identity verification result. After the verification passes, the user's basic information is obtained by matching the corresponding issued documents. At the same time, OCR processing is performed on the certificates provided by the user to extract text information and cross-verify it with the database information. The verified personal information and unique identifier are encrypted using the national cryptography algorithm SM4 to form a digital identity package. The user's basic information and unique application identification code are extracted from the digital identity package, and then, according to the user's destination country, the latest health entry requirements of that country are retrieved. The user's basic information and the destination country requirement data are input into a decision tree algorithm for analysis. This algorithm constructs a decision path based on the matching degree of the user's nationality, age, occupation, etc. with the destination country requirements and generates a personalized form structure. The system applies semantic analysis technology to the form structure to transform the abstract health requirements into specific form fields and pre-fill some fields according to the previous health records to reduce the user's filling burden. The information entered by the user in the interactive health form is checked in real time through data validity verification rules to ensure the rationality of the data format and content. The verified health declaration data is integrated with the digital identity package after being structured to generate a health data package with a digital signature.

[0016] The system extracts user information, destination country requirements, and health declaration information from health data packets to form a physical examination requirement set. This physical examination requirement set is input into the physical examination knowledge graph database for query. The knowledge inference engine analyzes the correlation between physical examination requirements and disease examination indicators in various countries around the world to generate an initial set of physical examination items. The machine learning algorithm analyzes the user's historical physical examination data, identifies health risk factors, adds targeted examinations to risk items, and forms a risk-weighted set of physical examination items. The system compares this set with the user's electronic health record, eliminates recently valid examination items, and obtains a set of physical examination items. This set is input into the shortest path algorithm, which considers the logical order of items and the layout of medical institutions to generate a physical examination path map. The system generates a QR code identifier for this path map, which contains the user identification code and the physical examination item code, and fuses the path map, item set, and QR code identifier with the health data packet to generate a physical examination execution package. The system extracts the user information, physical examination list, and path from the physical examination execution package, verifies the identity by scanning the QR code, and establishes a secure connection with distributed medical devices. When the devices collect blood analysis, imaging, and physiological index data, the system performs multi-modal data fusion, detects data integrity, and triggers a re-collection instruction for missing data. The heterogeneous original examination data is input into a multi-source heterogeneous data self-correction model, which consists of four-layer network structures: feature extraction, anomaly detection, data correction, and verification. After the examination data is processed by feature mapping, the system performs data consistency checks, calculates the covariance matrix and deviation threshold, and marks outliers. The Bayesian inference method combines historical physical examination data to calculate the conditional probability distribution, determines the type of anomaly, and performs corresponding corrections to obtain the corrected data. This data is converted into a medical standard format, and the whole process is recorded through the blockchain to form a structured physical examination result data set, which is associated with the user information to form a physical examination report package.

[0017] Extract the physical examination data and application information from the physical examination report package and input them into a multi-level review mechanism for analysis. This mechanism compares item by item the integrity of the physical examination items and the index range at the basic rule review layer. Subsequently, at the comprehensive analysis review layer, it integrates multiple examination results through machine learning algorithms to identify health risks. Finally, at the specific requirements review layer of the destination country, it conducts verification according to the conditions of the destination country. The system analyzes the abnormal situations and abnormal types of the review results at each layer, and if the prompt result of the abnormal situation is abnormal and does not meet the entry and exit requirements, it generates a review classification result. For the applications that pass the classification, the system generates an electronic health certificate containing an electronic signature and an anti-counterfeiting mark; for those that need to be rechecked, it generates an abnormal analysis report; for those that are rejected, it generates a non-compliance statement. All the review records are integrated with the classification results and processing documents to form a health certificate decision package. An electronic health certificate is generated according to the review results of the health certificate decision package. This certificate is digitally signed by PKI technology to generate a verification QR code. According to the receiving method selected by the user, the system performs multi-channel distribution. The electronic certificate is encrypted and sent to the email or application account, while the physical certificate sends a printing instruction to the self-service terminal. The system also performs cross-system data synchronization, pushes health information at different granularities to relevant government systems, establishes a certificate life cycle management, and provides a blockchain verification service, recording all operation logs to form a complete traceability chain.

[0018] In the embodiments of the present application, through the collection and processing of users' multimodal biometric information, combined with a deep neural network model for feature extraction, a unique identifier is generated, achieving high accuracy and security in identity verification and effectively preventing identity fraud. Based on the combination of digital identity packages and destination health requirement data, a dynamic form is generated through a decision tree algorithm, enabling the form content to accurately match the differentiated health requirements of different destinations, greatly improving the user experience and filling efficiency. By using a physical examination knowledge graph database in combination with a knowledge reasoning engine for intelligent matching of physical examination items, and through machine learning algorithms to identify health risk factors, a physical examination path map is generated in combination with the shortest path algorithm, effectively solving the problem of time waste caused by repeated examinations and unreasonable paths in traditional physical examinations. Especially in the aspect of medical data quality control, an innovative multi-source heterogeneous data self-correction model is adopted, which can identify outliers in examination data in real time. If there are abnormalities, it will prompt further reexamination. Through Bayesian inference methods, targeted processing is carried out on different types of outliers, significantly improving the accuracy and reliability of physical examination data. In the review link, a multi-level review mechanism is adopted in combination with a medical knowledge graph and an expert rule base, realizing the intelligence and high efficiency of the review process and reducing the subjectivity of manual review. Through PKI technology, digital signatures of electronic health certificates are carried out and distributed through multiple channels, associated with blockchain verification services, not only ensuring the authenticity and immutability of the certificates, but also providing a convenient online verification channel to meet the mutual recognition requirements of cross-border health certificates. Overall, the contributions of applying artificial intelligence algorithms and models in specific functional fields in the present invention are mainly reflected in: the deep neural network improves the accuracy of identity verification, the knowledge graph combined with the reasoning engine realizes intelligent matching of physical examination items, the multi-source heterogeneous data self-correction model solves the problem of medical data quality control, and the multi-level machine learning review mechanism replaces traditional manual review. The comprehensive application of these AI technologies enables the entire process of applying for an exit-entry health certificate to achieve digital transformation and intelligent upgrade, completely solving the problems of low efficiency, many human errors, data islands, and difficult cross-system verification in traditional methods.

[0019] In a specific embodiment, the process of executing step S101 may specifically include the following steps: Collect the face information, iris features, and fingerprint information of the user to obtain multimodal biometric data; Input the multimodal biometric data into a deep neural network model for feature extraction processing to generate a user unique identifier; Compare and verify the biometric templates stored in the security database based on the unique identifier to generate an identity verification result; According to the identity verification result, call a dedicated interface to connect to the pre-set exit-entry management database and health database to extract the user's basic information; Scan the passport or ID card provided by the user, extract the document text information through an improved OCR algorithm, cross-verify it with the user's basic information, and generate the verified personal information; Encrypt the verified personal information and the unique identifier through the national cryptographic algorithm SM4 to generate a digital identity package containing the user's basic information and the unique application identification code.

[0020] Specifically, obtain the user's face information, iris features, and fingerprint information through a high-resolution biometric collection device. For face information collection, an infrared and visible light dual-mode camera module is used to capture the three-dimensional face structure and texture features; for iris feature collection, a near-infrared light source is used to irradiate the iris to record unique texture patterns; for fingerprint information, the capacitive sensor is used to obtain the positional relationship between the fingerprint ridges and valleys. These three types of biometric information are digitally processed to form a multi-modal biometric data set containing pixel matrices, texture features, and ridge distribution maps. Input the collected multi-modal biometric data into a deep neural network model for feature extraction processing. The deep neural network model includes a convolutional layer, a pooling layer, and a fully connected layer. The convolutional layer is used to extract feature maps, the pooling layer performs feature dimensionality reduction, and the fully connected layer converts the features into feature vectors. For face information, the model extracts facial key points and texture features; for iris features, the model extracts circular texture encodings; for fingerprint information, the model extracts minutiae point positions and direction information. The features of each modality are fused through a feature fusion layer to generate a fixed-length feature code, which is transformed into a user unique identifier through a hash function.

[0021] Based on the generated unique identifier, the biometric templates stored in the security database are compared and verified. In the verification process, the feature distance calculation method is used to calculate the Euclidean distance or cosine similarity between the current feature and the database template. When the distance value is less than the preset threshold or the similarity is higher than the preset threshold, it is determined to be the same person and the verification passes; otherwise, the verification fails. The verification result includes the verification status (passed or failed) and the similarity score. After the verification passes, the dedicated interface is called according to the identity verification result to connect to the preset immigration management database and health database. The interface call uses an encrypted communication protocol to send a query request containing the unique identifier. The immigration management database returns basic identity information such as the user's name, gender, date of birth, nationality, passport number or ID number; the health database returns health data such as the user's previous medical examination records, vaccination status, and chronic medical history. These information are standardized in JSON format to form a user basic information dataset. Scanning the passport or ID certificate provided by the user is an important part of information supplementation and cross-verification. The certificate scanning uses a high-resolution image acquisition device to obtain the front and back images of the certificate. Through an improved OCR algorithm, which includes four steps: image preprocessing, text area detection, text recognition, and post-processing, the text information on the certificate is extracted. The improved OCR algorithm is optimized for the special characters and layouts of certificates from different countries, improving the recognition accuracy of special characters. The extracted certificate text information is cross-verified with the user basic information obtained from the database, comparing key fields such as name, certificate number, and date of birth to detect information consistency. After the verification passes, the information from each source is merged to generate the verified complete personal information. The verified personal information and the unique identifier are encrypted using the national secret algorithm SM4. SM4 is a Chinese commercial cipher algorithm that uses a 128-bit key and a symmetric block cipher system. The encryption process includes key expansion and round function transformation. In the encryption process, the personal information and the unique identifier are first serialized into a data stream and encrypted in 128-bit groups to generate ciphertext. The ciphertext is encapsulated together with the encryption metadata (encryption timestamp, encryption parameter identifier, etc.) to form a digital identity package. The digital identity package contains the encrypted user basic information and a unique application identification code derived from the unique identifier, which serves as a credential for identifying the user's identity in subsequent processes.

[0022] For example, user Mr. Wang needs to apply for a health certificate. When he comes to the self-service terminal, he aligns his face with the camera component, and the system captures his frontal face image and depth information. Subsequently, he gazes at the iris scanning area for a few seconds, and the system obtains his iris texture image. Finally, he places his right index finger on the fingerprint sensor, and the system captures the fingerprint image. These three sets of biometric data are input into a deep neural network, which respectively extracts a 128-dimensional face feature vector, a 256-dimensional iris feature vector, and a 96-dimensional fingerprint feature vector. After feature fusion, a 256-bit feature code is generated, and then a unique identifier is generated through the SHA-256 algorithm. The system compares the identifier with the database, calculates the similarity as 92%, which exceeds the verification threshold of 85%, and the verification is passed. Subsequently, the system calls the interface to obtain Mr. Wang's basic information, and at the same time scans his passport, and the OCR recognizes that the passport number is consistent with the database record. The system encrypts and processes through the SM4 algorithm to generate a digital identity package with anti-tampering characteristics.

[0023] In a specific embodiment, the process of executing step S102 may specifically include the following steps: Extract the user's basic information and the unique application identification code from the digital identity package to generate the user's basic information; According to the destination area where the user plans to go, obtain the health entry requirement data of the destination area; Input the user's basic information and the health entry requirement data of the destination area into a decision tree algorithm for analysis and processing to generate a personalized form structure; Apply semantic analysis to the personalized form structure, convert the health requirements into structured form fields, and perform pre-filling processing according to the user's previous health records to generate an interactive health form; Perform real-time verification processing on the information input by the user in the interactive health form through data validity verification rules to obtain formatted health declaration data; Perform structured processing on the formatted health declaration data and integrate it with the digital identity package, and generate a health data package through digital signature technology.

[0024] Specifically, the process of extracting the user's basic information and the unique application identification code from the digital identity package adopts decryption and deserialization technologies. Specifically, the SM4 decryption algorithm is used to decrypt the digital identity package to obtain the plaintext data corresponding to the ciphertext, and then the plaintext data is deserialized through a JSON parser to extract the user's basic information including fields such as name, gender, age, nationality, and previous health records, as well as the unique application identification code for identity identification. The extracted information is organized into structured user basic information, which serves as the data basis for subsequent form generation.

[0025] Obtain the health entry requirement data for the destination area where the user plans to travel; the decision tree algorithm constructs a multi-level tree structure, makes judgments based on specific attributes at each internal node, selects different branch paths according to the judgment results, and reaches a conclusion at the leaf node. In this method, the decision tree algorithm uses the user's nationality as the root node and selects different entry policy branches according to different nationalities; then uses the user's occupation as the second-level judgment node to subdivide the special requirements of different occupational groups; through this multi-level judgment process, a set of form fields that need to be filled in by this specific user is generated to form a personalized form structure. Apply semantic analysis technology to the personalized form structure, aiming to transform the abstract health requirements into specific form fields. The semantic analysis technology performs word segmentation and part-of-speech tagging on the health requirement text to identify the key entities and relationships therein; then matches the identified entities with the predefined form field templates through semantic matching to generate specific form field definitions, including attributes such as field name, data type, value range, and whether it is required. At the same time, pre-fill processing is performed according to the user's past health records. The system compares the form fields with the fields in the user's health records for matching, and automatically fills in the historical data for the successfully matched fields to reduce the user's filling burden. Through this processing, an interactive health form with pre-filled data is generated.

[0026] Performing real-time verification processing on the information entered by the user in the interactive health form through data validity verification rules is an important link to ensure data quality. The data validity verification rules include four types: format verification, range verification, logical verification, and consistency verification. Format verification checks whether the data conforms to the predefined format pattern, such as date format, email format, etc.; range verification checks whether the numerical data is within the logical range, such as height value, weight value, etc.; logical verification checks whether the logical relationship between multiple fields is reasonable, such as whether the disease history matches the gender; consistency verification checks whether there is a conflict between the information filled in by the user and the existing information in the system. The verification process adopts a real-time trigger mechanism, which verifies immediately when the user completes filling in a field, and immediately prompts and requires correction for the input that does not conform to the rules. Through this strict verification process, formatted health declaration data with correct format and reasonable content is obtained.

[0027] Structuring the formatted health declaration data and integrating it with the digital identity package, and generating a health data package through digital signature technology is the last step. The structuring process organizes various filled health declaration data according to a standardized data model, maps discrete form fields to predefined health data standards, and forms a dataset with a unified structure. Then, this structured data is integrated with the previous digital identity package to construct a user health application data body. Digital signature technology is used to sign the integrated data. Specifically, an asymmetric encryption algorithm is adopted, and the private key is used to encrypt the data digest to generate a digital signature, and the original data and the digital signature are encapsulated together to form a tamper-proof health data package.

[0028] In a specific embodiment, the process of executing step S103 may specifically include the following steps: Extract user identity information, destination area requirements, and health declaration information from the health data package to obtain a physical examination requirement set, and determine an initial physical examination item set according to the physical examination requirement set; Based on the initial physical examination item set, apply machine learning algorithms to analyze the user's historical physical examination data, identify health risk factors, and generate a risk-weighted physical examination item set; Compare and analyze the risk-weighted physical examination item set with the user's electronic health record, and eliminate the examination items that have been completed recently and have valid results to obtain a physical examination item set; Input the physical examination item set into the shortest path algorithm, consider the logical order between examination items and the spatial layout of medical institutions, and generate a physical examination path map; Generate a two-dimensional code identifier containing the user's unique application identification code and physical examination item codes for the physical examination path map, and integrate the physical examination path map, physical examination item set, two-dimensional code identifier with the health data package to generate a physical examination execution package.

[0029] Specifically, extract user identity information, destination area requirements, and health declaration information from the health data package. This extraction process uses data parsing and structured extraction technologies to deconstruct the health data package through a JSON parser or an XML parser, and extract the user's basic identity information (including name, gender, age, occupation category, etc.), specific health requirements of the destination area (including entry infectious disease screening requirements, occupational health requirements, etc.), and the health declaration information filled by the user (including past medical history, allergy history, medication history, etc.). The extracted information is organized into a physical examination requirement set according to a predefined data model, and each requirement item contains attributes such as requirement type, requirement source, and priority. Further, determine the initial physical examination item set according to the physical examination requirement set.

[0030] Based on the initial set of physical examination items, applying machine learning algorithms to analyze the user's historical physical examination data is to identify potential health risks. This process uses a supervised learning model that, by learning the association patterns between historical physical examination data and health risks, can predict potential risks based on the current user's historical physical examination results. The specific process extracts the user's historical physical examination data as feature vectors, including the historical values of various test indicators and their change trends; then inputs the feature vectors into a pre-trained risk prediction model, which uses algorithms such as random forest or gradient boosting tree to calculate the probability scores of various health risks; finally, based on the risk probability scores, the initial physical examination items are weighted, increasing the weights for the examination items associated with high risks to form a risk-weighted set of physical examination items, ensuring that key attention is paid to the user's health risk areas.

[0031] Comparing and analyzing the risk-weighted set of physical examination items with the user's electronic health record is to optimize the physical examination process. This process extracts the recently completed physical examination items and their results from the user's electronic health record, and the record includes the name of the examination item, the examination time, the examination result, and the validity period; then matches the name of each risk-weighted physical examination item with the items in the health record to find the potentially duplicate examination items; then judges the validity period of the successfully matched items, calculates the difference between the examination time and the current time, and compares it with the preset validity period of the item; finally, eliminates the examination items within the validity period with normal results from the risk-weighted set to obtain the set of physical examination items, avoiding unnecessary repeated examinations.

[0032] Inputting the set of physical examination items into the shortest path algorithm to generate a physical examination path map is the key to optimizing the user's physical examination process. The shortest path algorithm here considers two main factors: the logical order between examination items and the spatial layout of medical institutions. The logical order means that certain examinations must be carried out before and after other examinations. The algorithm constructs a weighted directed graph, where the nodes represent examination items and the edges represent the sequential relationship and spatial distance between items. The weights of the edges comprehensively consider the moving distance and the intensity of logical constraints. Then, an improved Dijkstra algorithm is applied to calculate the shortest path under all constraint conditions, generating a physical examination path map that not only conforms to the medical examination logic but also minimizes the user's moving distance within the medical institution.

[0033] Generating a QR code identifier for the physical examination path diagram and integrating relevant data to generate a physical examination execution package is the last step in preparing the physical examination process. During the QR code generation process, the user's unique application identification code and the physical examination item code are organized into a string according to a predetermined format; then the string is encoded into a two-dimensional matrix using the QR code encoding algorithm; finally, the matrix is rendered as an image to generate a scannable QR code identifier. The generation of the physical examination execution package is to fuse the physical examination path diagram, the set of physical examination items, the QR code identifier, and the original health data package, organize them into a structured data package, encapsulate it in the JSON or XML format, and add a digital signature to ensure data integrity. This physical examination execution package contains all the information required for the user to complete the physical examination.

[0034] In a specific embodiment, the process of executing step S104 may specifically include the following steps: Extract the user identity information, the physical examination item list, and the physical examination path from the physical examination execution package, perform identity verification by scanning the user's QR code identifier. When the verification is successful, establish a secure connection channel with the distributed medical device. When the verification fails, trigger a secondary biometric verification process, perform a comparison and analysis of the user's iris characteristics, generate a secondary verification result, and determine the authenticity of the identity based on the secondary verification result; Perform multi-modal data fusion on the blood analysis data, imaging data, and physiological index data collected by the distributed medical device to generate a heterogeneous raw examination data set; Input the heterogeneous raw examination data set into the multi-source heterogeneous data anomaly detection model. The multi-source heterogeneous data anomaly detection model is a four-layer neural network structure composed of a feature extraction layer, an anomaly detection layer, an anomaly marking layer, and a verification layer. Perform feature mapping processing on the examination data. When the feature dimension exceeds the preset threshold, perform a dimensionality reduction operation. When the dimension is lower than the threshold, keep the original dimension to obtain a set of feature vectors; Perform data consistency checking based on the set of feature vectors to generate an anomaly marking data set; Apply the Bayesian inference method to the anomaly marking data set for anomaly identification and analysis, calculate the conditional probability distribution in combination with historical physical examination data, judge the type of anomaly in the examination result. When it is a data outlier anomaly, generate a data anomaly prompt label. When it is a detection index anomaly, generate an index anomaly prompt label. When it is a normal physiological fluctuation, keep the original value without adding an anomaly label. Retain all the original examination data and add corresponding prompt labels to the anomaly items to obtain the original examination data with anomaly prompts; Convert the original inspection data with exception prompts into the standard format of medical data, record the timestamp of the entire collection process and operator information, count the items with exception prompts, and calculate the data integrity score. When the integrity score is higher than the high integrity threshold, directly generate a structured physical examination result dataset. When the integrity score is lower than the high integrity threshold, trigger the manual review process, integrate and generate a structured physical examination result dataset, and associate it with the user identity information to generate a physical examination report package.

[0035] Specifically, extract the user identity information, physical examination item list, and physical examination path from the physical examination execution package. This extraction process uses data parsing technology to deserialize the physical examination execution package, separating the identity data block containing the user's basic information, listing all the physical examination items that need to be completed, and the physical examination path information sorted by the optimal path. Subsequently, perform identity verification by scanning the user's QR code identifier. The QR code decoder reads the QR code content, extracts the user's unique application identification code and physical examination item code, and compares them with the identity information in the physical examination execution package. When the comparison result is consistent, the verification is successful, and an immediate secure connection channel to the distributed medical device is established, and encrypted communication is carried out using the TLS protocol to ensure the security of data transmission; when the comparison result is inconsistent, the verification fails, and the secondary biometric verification process is automatically triggered. Call the iris scanning device to collect the user's current iris image, extract the iris texture features, compare and analyze them with the pre-stored iris template, calculate the similarity score, generate the secondary verification result, and determine the authenticity of the user's identity based on this result.

[0036] After the identity verification is passed, start processing the inspection data collected by the distributed medical device. The distributed medical devices include blood analyzers, imaging devices, and physiological index monitoring devices, etc., which collect different types of physical examination data respectively. Blood analysis data includes numerical data such as blood cell count and biochemical indicators; imaging data includes image data such as X-ray films and ultrasounds; physiological index data includes measurement data such as blood pressure and heart rate. These data from different sources and different formats are integrated and processed through multimodal data fusion technology. Multimodal data fusion performs standardized preprocessing on various types of data, converting the original data output by different devices into a unified data format; then classifies and organizes the data according to the inspection item code, establishing the association relationship between the data; finally, organizes all the collected data into a structured heterogeneous original inspection data set.

[0037] Inputting the heterogeneous original inspection dataset into the multi-source heterogeneous data anomaly detection model is the core step of inspection data quality control. The multi-source heterogeneous data anomaly detection model is a deep learning model specifically used for medical data quality monitoring, consisting of four layers of neural network structures: the feature extraction layer, the anomaly detection layer, the anomaly marking layer, and the verification layer. The feature extraction layer applies different feature extraction algorithms to the input inspection data. For example, statistical feature extraction is performed on numerical data, convolutional feature extraction is performed on image data, and time-frequency feature extraction is performed on time series data to form feature representations. Then, the dimensions of the extracted features are evaluated. When the feature dimension exceeds the preset threshold, dimensionality reduction operations such as principal component analysis are performed to retain the main feature information; when the dimension is lower than the threshold, the original dimension is maintained without dimensionality reduction processing.

[0038] Performing data consistency checks based on the feature vector set is an important part of identifying abnormal data. The consistency check calculates the correlation between each data source, detects whether the logical relationship between the data is reasonable, then calculates the statistical distribution of each indicator, compares it with the normal reference range, and identifies the values that exceed the normal range; then calculates the continuity of the time series data to detect data jumps; finally, based on the above inspection results, each data point is scored, and the data points that exceed the threshold are marked as abnormal, generating an anomaly marking dataset containing anomaly types, anomaly degrees, and location information.

[0039] Applying the Bayesian inference method to the anomaly marking dataset for anomaly identification analysis is the key step to ensure data quality. The Bayesian inference method combines the user's historical physical examination data and the statistical distribution of the same type of population to calculate the conditional probability distribution, that is, the probability distribution of the current observed value given the historical data; then, based on the conditional probability distribution and the anomaly marking, the type of the anomaly value is judged. When it is judged as data outlier anomaly (for example, a certain indicator is too different from the user's historical value and does not conform to the physiological change law), a data anomaly prompt label is generated, and "data anomaly - suspected measurement error" is marked on the original data; when it is judged as a detection indicator anomaly (for example, key indicators such as blood pressure and blood sugar exceed the normal range; such as positive HIV test, positive hepatitis B surface antigen test, etc.), an indicator anomaly prompt label is generated, and "indicator anomaly - medical evaluation required" is marked on the original data; when it is judged as normal physiological fluctuation (for example, the detection value fluctuation caused by the user's physiological state changes such as taking medicine and eating), the original value is retained without adding an anomaly label. All original inspection data are completely retained, and only the corresponding prompt labels are added to the abnormal items to obtain the original inspection data with anomaly prompts.

[0040] Converting the original inspection data with anomaly prompts into the standard format of medical data is the final link in data integration. The standardization conversion adopts common data exchange standards in the medical field, such as HL7, DICOM, etc., to ensure data interoperability. The blockchain technology is embedded in the conversion process to record the whole process of data collection, including metadata such as the timestamp of each inspection item, operator identity authentication information, device identification, etc., to form a data traceability chain. The data integrity assessment calculates the integrity score based on preset assessment indicators, including data item coverage rate, key data missing rate, and the proportion of anomaly prompt items, etc. When the integrity score is higher than the high integrity threshold, a structured physical examination result dataset is directly generated; when the integrity score is lower than the high integrity threshold, an artificial review process is triggered, the data items that need to be manually reviewed are marked, and medical professionals are guided to conduct a recheck. Integrate all the original data with anomaly prompts, associate with the user identity information, and generate a formal physical examination report package.

[0041] In a specific embodiment, the process of executing step S105 may specifically include the following steps: Extract the user's physical examination data and application information from the physical examination report package to generate an audit input dataset; Input the audit input dataset into the first-layer basic rule audit layer of the multi-level audit mechanism composed of the medical knowledge graph and the expert rule library, conduct item-by-item comparison and analysis on the integrity of the physical examination items and the index range, and generate a basic audit result; Input the basic audit result into the second-layer comprehensive analysis audit layer of the multi-level audit mechanism, identify health risks by integrating multiple inspection results based on machine learning algorithms, conduct potential correlation analysis, and obtain a risk assessment result; Input the risk assessment result and the audit input dataset into the third-layer destination area specific requirement audit layer of the multi-level audit mechanism, conduct targeted verification according to the health entry conditions of the destination area, and generate a destination area compliance result; Calculate the credibility of the basic audit result, the risk assessment result, and the destination area compliance result. When the credibility of a certain audit result is lower than the credibility threshold, it is marked as an item for manual review, and an audit classification result is generated. The audit classification result classifies the application into three categories: passed, to be reviewed, and rejected; Generate corresponding processing documents based on the audit classification result. When the classification is passed, generate an electronic health certificate containing an electronic signature and an encrypted anti-counterfeiting mark. When the classification is to be reviewed, generate an anomaly analysis report. When the classification is rejected, generate a non-compliance statement, and integrate the audit classification result, the processing document, and the complete audit record to generate a health certificate decision package.

[0042] Specifically, the physical examination data and application information of the user are extracted from the physical examination report package. This extraction process uses data parsing technology to structurally read the physical examination report package, and respectively extracts the physical examination data containing the results of various examinations and the basic application information of the user. The extracted data undergoes format conversion and standardization processing, and is organized into a standardized data structure to form an audit input data set, which serves as the input for the multi-level audit mechanism. The basic rule audit layer is jointly composed of a medical knowledge graph and an expert rule library. The medical knowledge graph stores the logical relationships between physical examination items and medical reference ranges, while the expert rule library contains a large number of audit rules defined by medical experts. During the audit process, the integrity of the physical examination items is checked. By referring to the list of mandatory examination items required by the destination area, it is confirmed one by one whether they are completed. Then, the range comparison of each physical examination indicator is carried out, comparing the actual test value with the medical reference range, and marking the abnormal values that exceed the range. Next, the consistency between the examination items is checked to identify examination results with logical contradictions. The results of these checks are organized into a structured basic audit result, which includes information such as item completion status, indicator abnormality marking, and consistency assessment.

[0043] The comprehensive analysis audit layer is based on machine learning algorithms and integrates multiple examination results for in-depth analysis. This layer adopts an ensemble learning method, combines multiple classifiers and regression models, and comprehensively evaluates the user's physical examination data. During the processing, feature engineering is carried out on each examination indicator to extract derived features such as the correlation and trend of changes between indicators. Then, these features are input into a pre-trained health risk prediction model, which is trained based on a large amount of historical physical examination data and can identify potential health risk patterns. Next, abnormal pattern detection is performed to find combinations of indicators that match the characteristics of typical diseases. Finally, the prediction results of each model are summarized to generate a risk assessment result containing risk types, risk levels, and confidence levels.

[0044] The destination area specific requirements audit layer contains an entry health requirements database for countries and regions around the world, which stores the specific health requirements for different occupations and nationalities in different regions. During the audit process, according to the user's destination area and occupation category, the corresponding specific health requirements are queried. Then, these requirements are converted into structured verification rules. Next, by referring to the risk assessment result and the original physical examination data, the verification rules are executed item by item to check whether each requirement of the destination area is met. Finally, the verification results are summarized to generate a detailed destination area compliance result, clearly indicating the specific requirements that are met and not met.

[0045] Credibility calculations employ evidence theory, treating each level of audit results as an independent source of evidence and assessing the reliability of each conclusion. The calculation process assigns a basic credibility weight to each audit result, reflecting the importance of different audit levels in decision-making. Initial credibility is then adjusted based on data quality factors, such as the stability of test values and the calibration status of equipment. The internal consistency of the audit results is then analyzed, lowering credibility when conflicting conclusions arise between different levels. Finally, a comprehensive credibility score is calculated. When the credibility of a particular audit result falls below a preset credibility threshold, it is marked as requiring manual review. Based on the comprehensive audit results and credibility assessment, an audit classification result is generated, classifying applications into three categories: approved, requiring review, and rejected.

[0046] Generating the corresponding processing files based on the audit classification results is the final step in achieving the output of the audit results. For applications classified as approved, an electronic health certificate is generated. This certificate contains the user's basic information, a summary of the physical examination results, the validity period, and the certificate number, and adds an electronic signature and an encrypted anti-counterfeiting mark to ensure the authenticity and non-tamperability of the certificate; for applications classified as requiring review, a detailed abnormality analysis report is generated, clearly indicating the specific items that need to be reviewed, the reasons for the abnormality, and the review suggestions, to guide medical professionals to conduct targeted reviews; for applications classified as rejected, a non-conformity explanation document is generated, detailing the non-conformity items and reasons, and providing improvement suggestions. The audit classification results, corresponding processing files, and audit records are integrated together to form a health certificate decision package.

[0047] In a specific embodiment, the process of executing step S106 may specifically include the following steps: According to the review results in the health certificate decision package, the electronic health certificate information is identified, including user identity information, physical examination result summary, validity period and unique certificate identification code, to obtain the target data package; Apply PKI technology to digitally sign the target data packet and generate a verification QR code containing the core information of the certificate; Execute multi-channel distribution processing based on the user's pre-selected certificate receiving method. For users who choose electronic certificates, the encrypted certificate will be sent to the email and mobile application account. For users who choose physical certificates, the printing instruction will be sent to the self-service terminal device. Implement cross-system data synchronization based on the core certificate information in the electronic health certificate, push health information of different granularities to the management system through a secure API interface to form a synchronized data set; Set up a lifecycle management mechanism for electronic health certificates, establish a validity monitoring database, record certificate status information, and generate expiration reminder tasks; Associate the electronic health certificate with the blockchain verification service, provide an online verification channel through certificate QR code scanning or certificate number input, record logs for all certificate operations, and generate a certificate traceability chain.

[0048] Specifically, in the self-service registration method for exit-entry health certificates, multi-channel health certificate issuance and cross-system data synchronization are key links to ensure the effective circulation of health certificates. Identify the information of the electronic health certificate according to the review results in the health certificate decision package, adopt the templated certificate generation technology, extract the reviewed user identity information (including name, gender, date of birth, nationality, ID number, etc.), the summary of the physical examination results (including important health indicators and their result judgments), the expiration date (usually 12 months), and the unique certificate identification code automatically generated by the system (using a specific coding rule to ensure global uniqueness) from the decision package, and organize this information according to the international standard electronic certificate specification to generate a standardized target data packet. The target data packet is in XML or JSON format.

[0049] Applying PKI technology to the target data packet for digital signature processing is an important step to ensure the authenticity and integrity of the certificate. PKI (Public Key Infrastructure) technology is a security framework based on asymmetric encryption, used to implement digital signature and encrypted communication. The digital signature process calculates the hash value of the electronic health certificate, uses a secure hash algorithm such as SHA-256 to convert the certificate content into a fixed-length digest; then encrypts the digest with the private key of the issuing agency to generate a digital signature; finally, encapsulates the original certificate content and the digital signature together to form a signed electronic health certificate. At the same time, generate a verification QR code based on the core information of the certificate (such as certificate identification code, user identity information, expiration date, etc.), and this QR code adopts the QR code standard with a high error correction level to ensure the fault tolerance during printing and scanning, facilitating offline verification of the certificate authenticity.

[0050] Performing multi-channel distribution processing according to the certificate receiving method preselected by the user is an important link to achieve user-friendliness. The receiving method preselected by the user during the application process is recorded in the user preference settings, and the distribution system reads this setting to execute the corresponding distribution strategy. For users who choose the electronic certificate, the system encrypts and protects the signed certificate, encrypts the certificate content using a symmetric encryption algorithm (such as AES-256), and sends the key through the secure channel reserved by the user (such as mobile phone SMS); then sends the encrypted certificate file to the user's email through the secure email protocol, and at the same time pushes it to the user's personal account on the relevant mobile application to form a double backup. For users who choose the physical certificate, the system assembles the certificate data and printing parameters (such as the position of the anti-counterfeiting watermark, the layout of the holographic label, etc.) into a printing instruction, sends it to the specified self-service terminal device, triggers the printing process, and generates a physical certificate with physical anti-counterfeiting features.

[0051] Implementing a cross-system data synchronization mechanism based on the core certificate information in the electronic health certificate is a key step in achieving information interoperability. The data synchronization mechanism formulates a differentiated data push strategy according to the permission levels and data requirements of different receiving systems, determines the data granularity provided to each system, and then sends data packets to relevant government management systems through pre-configured secure API interfaces using the REST or SOAP protocol. The pushed content includes basic user identity information, certificate identification codes, physical examination summary information (with different levels of detail according to permission levels), expiration dates, and certificate statuses, etc. The scope of receiving systems includes the immigration management system, customs quarantine system, international health information exchange platform, etc. All pushed data is transmitted through a secure channel, and data verification is performed at the receiving end. After confirming integrity, it is written into the databases of each system to form a synchronized dataset.

[0052] Setting up a lifecycle management mechanism for the electronic health certificate is a necessary measure to ensure the timeliness of the health certificate. The lifecycle management establishes a dedicated expiration date monitoring database to record the issuance time, expiration date, and current status (valid, approaching expiration, expired, revoked, etc.) of each certificate. Then, it sets up a regular scanning task to scan the database daily to identify certificates approaching expiration (such as those expiring within 15 days). Next, it generates expiration reminder tasks for different reminder time points (such as 15 days, 7 days, and 3 days before expiration). Finally, it executes the reminder tasks and sends reminder messages through the notification channels pre-selected by users (such as text messages, emails, app push notifications, etc.) to inform users that the certificate is about to expire and the renewal method. In addition, for certificates that need to be revoked under special circumstances, the system immediately updates their status and synchronizes it to all relevant systems.

[0053] Associating the electronic health certificate with blockchain verification services is an innovative means of building a trusted verification system. The blockchain verification service, based on distributed ledger technology, provides an immutable record chain for certificate verification. The implementation process writes the key information of the certificate (such as certificate identification code, issuance time, expiration date, digital signature, etc.) into the blockchain to form a certificate record on the blockchain. Then, it provides multiple verification entrances, including verifying by scanning the QR code on the certificate or entering the certificate number through the official website. Next, the verification request triggers the execution of a smart contract to query the certificate record on the blockchain and perform cross-verification with the central database. Finally, it returns the verification result, including the authenticity, current status, and basic information of the certificate. At the same time, all certificate operations (such as issuance, verification, revocation, etc.) are automatically logged in detail, including the operation time, operation type, operator, and operation content, etc. These operation logs are also written into the blockchain to form a certificate traceability chain, ensuring the transparency and auditability of the entire lifecycle of the certificate.

[0054] The above describes the self-service registration method for the exit-entry health certificate in the embodiments of the present application. Next, the self-service registration system for the exit-entry health certificate in the embodiments of the present application will be described. Please refer to Figure 2 , an embodiment of the self-service registration system for the exit-entry health certificate in the embodiments of the present application includes: A collection module 201, configured to collect and process the multimodal biometric information of a user to generate a digital identity package; An integration module 202, configured to perform dynamic form generation processing according to the user's basic information and destination area health requirement data in the digital identity package, obtain health declaration information, and integrate the health declaration information with the digital identity package to generate a health data package; A query module 203, configured to query and process the physical examination knowledge graph database based on the health data package, match the physical examination item set and generate a physical examination path map, and fuse the physical examination path map with the health data package to generate a physical examination execution package; A control module 204, configured to perform real-time quality control processing on the inspection data collected by the distributed medical devices according to the physical examination execution package, identify outliers in the inspection data by using a multi-source heterogeneous data self-correction model, generate a structured physical examination result data set, and generate a physical examination report package; An analysis module 205, configured to input the physical examination report package into a multi-level review mechanism for compliance analysis processing, generate a review result and an electronic health certificate, and generate a health certificate decision package; A synchronization module 206, configured to perform certificate issuance and cross-system data synchronization processing based on the health certificate decision package, and implement multi-channel distribution and verification of the electronic health certificate.

[0055] Through the collaborative cooperation of the above-mentioned various components, by collecting and processing the user's multimodal biometric information, extracting features in combination with a deep neural network model, generating a unique identifier, high accuracy and security of identity verification are achieved, effectively preventing identity fraud; based on the combination of the digital identity package and the health requirement data of the destination area, dynamic form generation is carried out through the decision tree algorithm, so that the form content accurately matches the different health requirements of different destinations, greatly improving the user experience and filling efficiency; the physical examination knowledge graph database is used in combination with the knowledge reasoning engine to perform intelligent matching of physical examination items, and machine learning algorithms are used to identify health risk factors, and the physical examination path map is generated in combination with the shortest path algorithm, effectively solving the problem of time waste caused by repeated examinations and unreasonable paths in traditional physical examinations; especially in the aspect of medical data quality control, an innovative multi-source heterogeneous data self-correction model is adopted, which can identify outliers in the inspection data in real time. If there are abnormalities, it will prompt further reexamination. Targeted processing of different types of outliers is carried out through Bayesian inference methods, significantly improving the accuracy and reliability of physical examination data; in the review link, a multi-level review mechanism is adopted in combination with the medical knowledge graph and the expert rule base, realizing the intelligence and high efficiency of the review process, and reducing the subjectivity of manual review; through PKI technology, digital signatures of electronic health certificates are carried out and distributed through multiple channels, and are associated with blockchain verification services, which not only ensures the authenticity and immutability of the certificates, but also provides a convenient online verification channel, meeting the mutual recognition requirements of cross-border health certificates; overall, the contribution of applying artificial intelligence algorithms and models in the specific functional field of the present invention is mainly reflected in: the deep neural network improves the accuracy of identity verification, the knowledge graph cooperates with the reasoning engine to achieve intelligent matching of physical examination items, the multi-source heterogeneous data self-correction model solves the problem of medical data quality control, and the multi-level machine learning review mechanism replaces the traditional manual review. The comprehensive application of these AI technologies has enabled the entire process of applying for an entry-exit health certificate to achieve digital transformation and intelligent upgrading, completely solving the problems of low efficiency, many human errors, data islands, and difficult cross-system verification in traditional methods.

[0056] Above Figure 2 From the perspective of modular functional entities, the self-service registration system for entry-exit health certificates in the embodiments of the present invention is described in detail. Below, the self-service registration device for entry-exit health certificates in the embodiments of the present invention is described in detail from the perspective of hardware processing.

[0057] Figure 3FIG. 0 is a schematic structural diagram of a self-service registration device for an entry-exit health certificate provided by an embodiment of the present invention. The self-service registration device 300 for an entry-exit health certificate may vary greatly due to different configurations or performances, and may include one or more processors (central processing units, CPUs) 310 (for example, one or more processors) and a memory 320, and one or more storage media 330 for storing application programs 333 or data 332 (for example, one or more mass storage device terminals). Among them, the memory 320 and the storage media 330 may be transient storage or persistent storage. The program stored in the storage media 330 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations on the self-service registration device 300 for an entry-exit health certificate. Further, the processor 310 may be configured to communicate with the storage media 330 and execute a series of instruction operations in the storage media 330 on the self-service registration device 300 to implement the steps of the above-mentioned self-service registration method for an entry-exit health certificate.

[0058] The self-service registration device 300 for an entry-exit health certificate may further include one or more power supplies 340, one or more wired or wireless network interfaces 350, one or more input / output interfaces 360, and / or one or more operating systems 331, such as Windows Serve, Mac OS X, Unix, Linux, FreeBSD, etc. Those skilled in the art can understand that Figure 3 The shown structural diagram of the self-service registration device for an entry-exit health certificate does not limit the self-service registration device for an entry-exit health certificate provided by the present invention, and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0059] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions run on a computer, the computer is caused to execute the steps of the self-service registration method for an entry-exit health certificate.

[0060] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the above-described systems, systems, and units can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0061] When the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions to cause a self-service registration device for the entry-exit health certificate (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs.

[0062] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A self-registration method for an entry-exit health certificate, characterized in that, The method includes: Collect and process the multi-modal biometric information of the user to generate a digital identity package; Perform dynamic form generation processing according to the user's basic information in the digital identity package and the health requirement data of the destination area, obtain the health declaration information, and integrate the health declaration information with the digital identity package to generate a health data package; Query and process the physical examination knowledge graph database based on the health data package, match the physical examination item set and generate a physical examination path map, and fuse the physical examination path map with the health data package to generate a physical examination execution package; Perform real-time quality control processing on the examination data collected by the distributed medical devices according to the physical examination execution package, identify outliers in the examination data using a multi-source heterogeneous data self-correction model, generate a structured physical examination result data set, and generate a physical examination report package; Input the physical examination report package into a multi-level review mechanism for compliance analysis processing, generate a review result and an electronic health certificate, and generate a health certificate decision package; Perform certificate issuance and cross-system data synchronization processing based on the health certificate decision package to achieve multi-channel distribution and verification of the electronic health certificate.

2. The self-registration method for the entry-exit health certificate according to claim 1, wherein, The collection and processing of the multi-modal biometric information of the user to generate a digital identity package includes: Collect the user's face information, iris features and fingerprint information to obtain multi-modal biometric data; Input the multi-modal biometric data into a deep neural network model for feature extraction processing to generate a user unique identifier; Compare and verify the biometric template stored in the security database based on the unique identifier to generate an identity verification result; Call a dedicated interface according to the identity verification result to connect to the pre-set entry-exit management database and health database, and extract the user's basic information; Scan the passport or ID card provided by the user, extract the document text information through an improved OCR algorithm, and cross-verify it with the user's basic information to generate verified personal information; Encrypt the verified personal information and the unique identifier through the national cryptography algorithm SM4 to generate a digital identity package containing the user's basic information and a unique application identification code.

3. The self-registration method for the entry-exit health certificate according to claim 1, characterized in that, The dynamic form generation processing according to the user's basic information in the digital identity package and the health requirement data of the destination area, obtaining the health declaration information, and integrating the health declaration information with the digital identity package to generate a health data package includes: Extract the user's basic information and a unique application identification code from the digital identity package to generate user basic information; Obtain the health entry requirement data of the destination area according to the destination area where the user plans to go; Input the user's basic information and the health entry requirement data of the destination area into a decision tree algorithm for analysis processing to generate a personalized form structure; Apply semantic analysis to the personalized form structure, convert the health requirements into structured form fields, and perform pre-filling processing according to the user's past health records to generate an interactive health form; Perform real-time verification processing on the information input by the user in the interactive health form through data validity verification rules to obtain formatted health declaration data; The formatted health declaration data is structured and integrated with the digital identity package, and a health data package is generated through digital signature technology.

4. The self-registration method for the entry-exit health certificate according to claim 1, characterized in that, The query processing of the physical examination knowledge graph database based on the health data packet, matching the physical examination item set and generating a physical examination path map, and fusing the physical examination path map with the health data packet to generate a physical examination execution package includes: Extracting user identity information, destination area requirements, and health declaration information from the health data packet to obtain a physical examination requirement set, and determining an initial physical examination item set based on the physical examination requirement set; Based on the initial set of physical examination items, applying a machine learning algorithm to analyze the user's historical physical examination data, identify health risk factors, and generate a risk-weighted physical examination item set; Comparing and analyzing the risk-weighted physical examination item set with the user's electronic health record, eliminating recently completed examination items with valid results, and obtaining a physical examination item set; Input the set of physical examination items into the shortest path algorithm, consider the logical order between the examination items and the spatial layout of the medical institutions, and generate a physical examination path map; A QR code identifier containing a user's unique application identification code and a physical examination item code is generated for the physical examination roadmap, and the physical examination roadmap, the physical examination item set, the QR code identifier and the health data packet are integrated to generate a physical examination execution package.

5. The self-registration method for the exit-entry health certificate according to claim 1, characterized in that, The physical examination execution package performs real-time quality control processing on the inspection data collected by the distributed medical equipment, uses a multi-source heterogeneous data self-correction model to identify outliers in the inspection data, generates a structured physical examination result data set, and generates a physical examination report package, including: Extract user identity information, a list of physical examination items, and a physical examination path from the physical examination execution package, perform identity verification by scanning the user's QR code, establish a secure connection channel with the distributed medical device when the verification is successful, and trigger a secondary biometric verification process when the verification fails, compare and analyze the user's iris features, generate a secondary verification result, and determine the authenticity of the identity based on the secondary verification result; Performing multimodal data fusion on the blood analysis data, imaging data, and physiological indicator data collected by the distributed medical equipment to generate a heterogeneous original examination data set; Inputting the heterogeneous original inspection data set into the multi-source heterogeneous data anomaly detection model, the multi-source heterogeneous data anomaly detection model is a four-layer neural network structure consisting of a feature extraction layer, an anomaly detection layer, an anomaly marking layer, and a verification layer. Feature mapping is performed on the inspection data. When the feature dimension exceeds a preset threshold, a dimensionality reduction operation is performed. When the dimension is below the threshold, the original dimension is maintained to obtain a feature vector set; Performing data consistency check based on the feature vector set to generate an anomaly labeled data set; Apply the Bayesian inference method to the abnormal marked data set for abnormal recognition and analysis, calculate the conditional probability distribution in combination with historical physical examination data, judge the type of abnormal examination results, generate a data anomaly prompt label when it is a data outlier anomaly, generate an index anomaly prompt label when it is a detection index anomaly, and retain the original value without adding an anomaly label when it is a normal physiological fluctuation. Retain all the original examination data and add corresponding prompt labels to the abnormal items to obtain the original examination data with anomaly prompts. Convert the original examination data with anomaly prompts into the standard format of medical data, record the time stamps of the entire collection process and the operator information, count the anomaly prompt items and calculate the data integrity score. When the integrity score is higher than the high integrity threshold, directly generate a structured physical examination result data set. When the integrity score is lower than the high integrity threshold, trigger the manual review process, integrate and generate a structured physical examination result data set, and associate it with the user identity information to generate a physical examination report package.

6. The self-registration method for the entry-exit health certificate according to claim 1, characterized in that Input the physical examination report package into a multi-level review mechanism for compliance analysis and processing, generate a review result and an electronic health certificate, and generate a health certificate decision package, including: Extract the physical examination data and application information of the user from the physical examination report package to generate a review input data set; Input the review input data set into the first-level basic rule review layer of the multi-level review mechanism composed of a medical knowledge graph and an expert rule library, conduct item-by-item comparison and analysis on the integrity of the physical examination items and the index range, and generate a basic review result; Input the basic review result into the second-level comprehensive analysis review layer of the multi-level review mechanism, identify health risks by integrating multiple examination results based on machine learning algorithms, and conduct potential correlation analysis to obtain a risk assessment result; Input the risk assessment result and the review input data set into the third-level destination area specific requirement review layer of the multi-level review mechanism, conduct targeted verification according to the health entry conditions of the destination area, and generate a destination area compliance result; Calculate the credibility of the basic review result, the risk assessment result and the destination area compliance result. When the credibility of a certain review result is lower than the credibility threshold, mark it as a manual review item and generate a review classification result. The review classification result classifies the application into three categories: passed, to be reviewed, and rejected; Generate corresponding processing documents based on the review classification result. When the classification is passed, generate an electronic health certificate containing an electronic signature and an encrypted anti-counterfeiting identifier. When the classification is to be reviewed, generate an anomaly analysis report. When the classification is rejected, generate a non-compliance statement, and integrate the review classification result, the processing document and the complete review record to generate a health certificate decision package.

7. The self-registration method for the exit-entry health certificate according to claim 1, wherein Perform certificate issuance and cross-system data synchronization processing based on the health certificate decision package to achieve multi-channel distribution and verification of the electronic health certificate, including: Identify the information of the electronic health certificate according to the review result in the health certificate decision package, including user identity information, a summary of the physical examination results, the expiration date and a unique certificate identification code, to obtain a target data package; Apply PKI technology to digitally sign the target data package and generate a verification QR code containing the core information of the certificate; Execute multi-channel distribution processing according to the certificate receiving method preselected by the user. For users who select e-certificates, send the encrypted certificate to the email and mobile application accounts. For users who select physical certificates, send the printing instructions to the self-service terminal device; Execute a cross-system data synchronization mechanism based on the core certificate information of the e-health certificate, and push health information at different granularities to the management system through a secure API interface to form a synchronized data set; Set a life cycle management mechanism for the e-health certificate, establish a validity monitoring database, record the certificate status information, and generate an expiration reminder task; Associate the e-health certificate with the blockchain verification service, provide an online verification channel through certificate QR code scanning or certificate number input, record logs for all certificate operations, and generate a certificate traceability chain.

8. A self-service registration system for entry-exit health certificates, characterized in that, For implementing the self-service registration method for the entry-exit health certificate as described in any one of claims 1-7, the self-service registration system for the entry-exit health certificate includes: A collection module for collecting and processing the multi-modal biometric information of the user to generate a digital identity package; An integration module for performing dynamic form generation processing based on the user's basic information and destination area health requirement data in the digital identity package, obtaining health declaration information, and integrating the health declaration information with the digital identity package to generate a health data package; A query module for querying the physical examination knowledge graph database based on the health data package, matching the physical examination item set and generating a physical examination path map, and fusing the physical examination path map with the health data package to generate a physical examination execution package; A control module for performing real-time quality control processing on the inspection data collected by distributed medical devices according to the physical examination execution package, using a multi-source heterogeneous data self-correction model to identify outliers in the inspection data, generating a structured physical examination result data set, and generating a physical examination report package; An analysis module for inputting the physical examination report package into a multi-level review mechanism for compliance analysis processing, generating a review result and an e-health certificate, and generating a health certificate decision package; A synchronization module for performing certificate issuance and cross-system data synchronization processing based on the health certificate decision package to achieve multi-channel distribution and verification of the e-health certificate.

9. A self-service registration device for an entry-exit health certificate, characterized in that, It includes a memory and a processor. The memory stores a computer program that can run on the processor. When the processor executes the computer program, it implements the self-service registration method for the entry-exit health certificate as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is run by the processor, it causes the processor to execute the self-service registration method for the entry-exit health certificate as described in any one of claims 1 to 7.

Images