Ophthalmology department multi-source heterogeneous data fusion storage and cross-platform transmission method and device, computer equipment and storage medium

Abstract

The invention discloses an ophthalmology multi-source heterogeneous data fusion storage and cross-platform transmission method and device, computer equipment and a storage medium, and relates to the technical field of ophthalmology data processing. According to the ophthalmology department multi-source heterogeneous data fusion storage and cross-platform transmission method and device, the computer equipment and the storage medium, data of different structures can be collected from various ophthalmology department data sources, a storage model and an interface are constructed, and ophthalmology department multi-source heterogeneous data fusion storage is achieved; seamless data transmission among different operating systems and various devices can be ensured; the ophthalmology data can be encrypted, access authority is set, and backup recovery operation is executed; data traffic and quality can be monitored in real time, and system logs can be effectively managed. The system can be integrated with a related big data analysis platform and an enterprise informatization system, solves the problems in the aspects of ophthalmology multi-source heterogeneous data fusion storage and cross-platform transmission, and practically improves the comprehensive utilization value of ophthalmology data.

Description

Technical Field

[0001] The present invention relates to the technical field of ophthalmic data processing, and in particular to a method, device, computer equipment and storage medium for fusion storage and cross-platform transmission of ophthalmic multi-source heterogeneous data. Background Art

[0002] In the field of ophthalmic medical treatment and research, data sources are becoming increasingly extensive and their structures are complex and diverse. They include structured patient basic information database data, structured test data generated by ophthalmic examination equipment, such as intraocular pressure measurement data, vision test data, etc., unstructured ophthalmic medical record document data, such as doctors' diagnosis records, surgical records, etc., and multimedia data, such as fundus images, OCT optical coherence tomography images, FFA fundus fluorescein angiography image sequences, UBM ultrasound biomicroscope examination images and videos, etc. These data have significant differences in structure and format due to different sources, which brings huge challenges to data management, storage and sharing.

[0003] Traditional data storage and transmission methods are mostly designed for a single data type or a specific platform, which exposes many problems in ophthalmic data processing. For example, it is difficult for structured databases to effectively store and manage unstructured medical records and multimedia ophthalmic image and video data; data formats are incompatible and transmission protocols are inconsistent between different operating systems (such as Windows server systems within hospitals, iOS or Android mobile terminal systems used by doctors, etc.) and device types (such as desktop examination equipment, portable examination instruments, servers, personal computers, mobile terminals, etc.), making it extremely difficult to share and interact with ophthalmic data on different platforms, which seriously limits the comprehensive utilization value of ophthalmic data and cannot meet the needs of ophthalmic precision diagnosis, telemedicine collaboration, and scientific research analysis for data processing.

[0004] In this context, the present invention provides a method, apparatus, computer equipment and storage medium for fusion storage and cross-platform transmission of multi-source heterogeneous ophthalmic data. Summary of the invention

[0005] 1. Technical issues to be solved

[0006] The purpose of the present invention is to provide a method, device, computer equipment and storage medium for fusion storage and cross-platform transmission of multi-source heterogeneous ophthalmic data, so as to solve the problems of difficulty in fusion storage and inconvenience in cross-platform transmission of multi-source heterogeneous ophthalmic data in the prior art, and to improve the comprehensive utilization value of ophthalmic data.

[0007] (II) Technical solution

[0008] The object of the present invention is to provide a method, device, computer equipment and storage medium for fusion storage and cross-platform transmission of multi-source heterogeneous ophthalmic data, including:

[0009] Data collection module, as the key entry point for ophthalmic data inflow, has the ability to identify data sources and extract data. Its data source identification submodule uses intelligent algorithms and feature matching technology to quickly and accurately identify the source of data, whether it is the mainstream database system MySQL and Oracle used to store basic patient information and some inspection equipment data, covering the file formats under Windows and Linux operating systems, including file systems of various types of ophthalmic medical records, including web data, cloud storage data, various network data sources involving data sharing in telemedicine collaboration or scientific research data acquisition, or specific application program interfaces such as ophthalmic examination software interfaces and medical record management system interfaces customized within the hospital. For example, when identifying the database source, through deep analysis of port numbers, connection strings, and database-specific system table information features, the database type and version information can be determined in a very short millisecond time; for the file system, it can not only determine the common file format based on the file header information, but also identify the complex file storage system through the file directory structure and metadata information, such as accurately identifying the classification storage structure and related metadata information of ophthalmic medical records.

[0010] The data extraction submodule applies targeted extraction strategies based on the recognition results. For structured database data, the target data can be accurately extracted through carefully constructed SQL query statements or efficient database connection tools: JDBC, ODBC. When processing large-scale structured data, the database connection pool technology and optimized query execution plan are used to improve the extraction efficiency compared with traditional extraction methods. For example, when extracting the intraocular pressure measurement history data of a large number of patients, the required information can be obtained quickly and accurately. For unstructured document data: ophthalmic medical records in Word, PDF, and TXT formats, regular expression-based and DOM parsing are enabled to deeply mine text content. Taking the processing of large PDF medical records containing rich examination results and diagnostic analysis as an example, the multi-threaded parallel parsing technology can shorten the parsing time and efficiently extract the key information. For multimedia data: fundus images, OCT images, FFA videos, UBM images and videos, professional multimedia data reading libraries are used: OpenCV processes images and FFmpeg processes audio and video to completely extract data information. In image data extraction, OpenCV's efficient pixel processing algorithm is used to obtain detailed pixel information, color features, and size of images in a very short time. The processing speed is improved compared to ordinary image reading libraries. For example, the pixel distribution and morphological features of blood vessels in fundus images can be quickly obtained. For video data, key data such as the video frame rate, resolution, encoding format, and key frame information in the image sequence can be accurately extracted to provide comprehensive data support for subsequent disease analysis and diagnosis.

[0011] Data fusion storage module: The metadata description layer of this module uses a relational database table structure to store metadata information of multi-source heterogeneous ophthalmic data. These metadata are like the "identity cards" of the data, recording in detail key elements such as the data source (such as a certain model of inspection equipment from a hospital's ophthalmology clinic, a certain doctor's specific medical record, etc.), type (structured, unstructured, multimedia), format (such as database table structure, Word document format, JPEG image format, etc.), creation time, and association relationship (such as the patient's fundus image is closely related to the corresponding medical record document, and the OCT image and intraocular pressure examination data have an inherent logical relationship).

[0012] The data storage layer stores data in categories based on data characteristics. Structured data is stored using mature relational database storage engines (such as InnoDB, MyISAM, etc.). In data insertion and query operations, the index structure is optimized by using B+ tree indexes combined with index coverage technology. Unstructured document data is stored in distributed file systems Ceph and GlusterFS. The distributed storage and high fault tolerance characteristics of distributed file systems are used to ensure data reliability and scalability. Even if some storage nodes fail, the integrity and availability of medical record document data can be guaranteed. Multimedia data is based on the MinIO multimedia repository of object storage, and a close mapping relationship with the metadata description layer is established through the designed index mechanism to ensure the efficiency of data query and positioning.

[0013] In addition, the data index construction submodule uses a combination of inverted index and B-tree index to build a multi-level index system based on data keywords, metadata information or data content features, which greatly improves the data retrieval speed. For the keyword index construction of structured data, a dynamic index update strategy is adopted; for the multimedia data content feature index, a deep learning algorithm is used to extract the lesion feature vector of the fundus image and the retinal layer feature vector of the OCT image to build a feature index. In the retrieval of similar lesion images and related examination images, the accuracy is improved, providing doctors with a powerful technical means to quickly compare and analyze the changes in patients' conditions and similar cases.

[0014] The data transmission module is designed from multiple dimensions to achieve seamless transmission across platforms. The platform detection submodule uses a combination of system detection technology and network scanning technology to comprehensively detect the operating system types of the source and target platforms of data transmission, such as Windows systems for hospital servers and iOS or Android systems for doctors' mobile terminals, and detects device types, such as large ophthalmic examination equipment, desktop computers, laptops, tablet computers, and smartphones, as well as network environment information, including bandwidth, latency, and packet loss rate.

[0015] The transport protocol adapter module selects the adapted protocol from the TCP, UDP, HTTP, and FTP transport protocol libraries based on the test results. In scenarios with low bandwidth and high real-time requirements, such as when doctors use mobile terminals to view fundus images or videos of patients during remote consultations, the lightweight UDP protocol is preferred, and the custom protocol of the application layer is combined to optimize the transmission efficiency; in environments with high bandwidth and strict requirements on data integrity, such as when a large amount of ophthalmic examination data and medical records are transmitted between internal hospital systems, the reliable TCP protocol is selected, and SSL / TLS encrypted transmission is used to ensure data security and prevent data from being stolen or tampered with during transmission.

[0016] The data format conversion submodule builds a dynamic format recognition and conversion mechanism for different ophthalmic data types, including structured, unstructured, images, and videos. When the system recognizes ophthalmic image data, it detects the image content through an image recognition algorithm based on deep learning, thereby automatically determining the best image compression and format conversion strategy; it proposes intelligent format conversion based on deep learning and image content, which can perform targeted compression according to the lesion area in the ophthalmic image to avoid the loss of important diagnostic information during the compression process; it enhances the intelligence of the image processing process and improves the quality and efficiency of the image during transmission;

[0017] The process of building a dynamic format recognition and conversion mechanism includes: using the data acquisition module to identify and classify the source of the input ophthalmic data to ensure that data from different sources can be correctly classified and processed. The identified data types include but are not limited to structured data, unstructured document data, image data, and video data; after the data is classified, the system needs to further identify the characteristics of each data format, and select the most appropriate conversion method and format for the types of image, video, and document data. According to the types of identified images, videos, or documents and the requirements of the target device, the system will perform corresponding compression and conversion to achieve the best transmission efficiency and user experience. In order to ensure the real-time format conversion and the stability of transmission, the system should design a real-time monitoring and feedback mechanism to promptly handle network fluctuations and data format incompatibility issues. The transmission requests of images, videos, and documents are used as input, and the network bandwidth, delay, and data traffic are monitored in real time. When the system detects that the network bandwidth is insufficient or the delay is too high, it automatically adjusts the transmission strategy, outputs the optimized image, video, and document data transmission stream, and transmits the converted and optimized data to the target device or platform.

[0018] The data security management module protects data security in all aspects. The data encryption submodule integrates the AES algorithm in the symmetric encryption algorithm and the RSA algorithm in the asymmetric encryption algorithm to perform layered encryption for sensitive ophthalmic data. First, a large amount of ophthalmic examination data and medical records are quickly encrypted using a symmetric encryption algorithm, and then an asymmetric encryption algorithm is used to encrypt and transmit the symmetric encryption key to ensure data confidentiality and key security.

[0019] The 256-bit encryption key length of the AES algorithm is used, and the encryption strength is increased by 2^128 times compared to the 128-bit key. In the data encryption performance test, the encryption speed can reach more than 100MB per second, which can not only ensure data security, but also meet the needs of rapid processing of ophthalmic data.

[0020] The access control submodule is based on the role-based access control RBAC model. According to the user role (such as ophthalmologists, general ophthalmologists, nurses, researchers, patients, etc.) and data sensitivity (such as patient privacy information, research data of rare ophthalmic diseases, etc.), fine access rights are set at the data storage and access level, covering read, write, modify and delete permissions, and only users or systems with corresponding permissions are allowed to access specific ophthalmic data. For example, ophthalmologists can view and modify the detailed medical records and examination data of all patients in order to make comprehensive diagnosis and formulate treatment plans; general ophthalmologists can only view and modify part of the data of the patients they are responsible for to ensure the privacy and security of patient data; nurses can only view the basic information of patients and nursing-related data for daily nursing work; researchers can access anonymized ophthalmic data for research after obtaining authorization to promote the progress of ophthalmic medical research; patients can only view their basic examination results and diagnostic information so that they can understand their own condition. Through the audit function of the permission management system, permission operations are recorded in detail to trace and supervise data access behavior and prevent data leakage and illegal access.

[0021] The data backup and recovery submodule adopts a strategy that combines full backup and incremental backup. It regularly performs full backup of stored ophthalmic data and incremental backup when the data is updated. The backup cycle is flexibly set according to the importance and update frequency of the data. Newly generated inspection data is incrementally backed up daily, and all data is fully backed up weekly. When data is lost or damaged, the data is accurately restored to the specified time point with the help of the recovery algorithm of backup data and data log-based recovery technology to ensure data availability and integrity. In the data recovery test, the average recovery time for ophthalmic data recovery requests at different time points can be controlled within 5 minutes, and the integrity and accuracy verification pass rate of the restored data is 100%, ensuring that ophthalmic medical work and scientific research analysis are not affected by data loss. For example, data can be quickly restored after a system failure or accidental deletion of data to ensure the normal operation of the hospital's ophthalmic business.

[0022] The data monitoring and management module closely monitors and meticulously manages the entire process of ophthalmic data processing. The data flow monitoring submodule uses network flow monitoring technology and data flow statistical algorithms to monitor the flow of data in real time during storage and transmission, accurately count the inflow, outflow and data transmission rate, and based on the preset threshold curve based on historical flow data and ophthalmic data processing system performance indicators, that is, the flow threshold model, when the data flow exceeds the threshold, an alarm is issued in time to quickly locate and solve the abnormal flow problem. In the process of flow monitoring, the real-time analysis algorithm of flow data is used, which can issue an alarm within 1 second after the flow abnormality occurs, and the false alarm rate is controlled below 2%, ensuring the stability and reliability of ophthalmic data transmission, avoiding data transmission delays or interruptions caused by flow problems, and affecting the timeliness of ophthalmic diagnosis and treatment.

[0023] The data quality monitoring submodule uses a data verification rule library and a data cleaning and repair algorithm to conduct in-depth monitoring of the quality of stored ophthalmic data, and comprehensively check the integrity of the data (such as whether the required fields in the patient information form are complete, whether there are missing ophthalmic examination data records, etc.), accuracy (such as whether the intraocular pressure measurement data is within a reasonable range, whether the diagnostic information in the medical record is accurate and standardized, etc.) and consistency (such as whether the vision data of the same patient generated by different examination equipment is consistent, etc.). For data that does not meet the quality standards, the corresponding processing flow is automatically triggered according to the type of data problem, including data cleaning (such as removing duplicate ophthalmic examination records, correcting erroneous intraocular pressure data, etc.), data repair (such as repairing missing medical record information based on data association relationships, etc.) or data deletion (such as serious errors or expired examination data) to ensure that the data quality meets the needs of ophthalmic medical care and scientific research. In the data quality monitoring test, for large-scale ophthalmic data samples, the accuracy of data integrity verification can reach more than 99%, the accuracy of accuracy verification can reach more than 98%, and the accuracy of consistency verification can reach more than 97%, providing a reliable data foundation for accurate ophthalmic diagnosis and scientific research, such as ensuring the accuracy and reliability of data in scientific research analysis. The data verification rule library contains integrity verification, accuracy verification, and consistency verification rules. Based on the data cleaning and repair algorithm, the duplicate ophthalmic data is cleaned by using a hash algorithm-based duplicate checking technology, and the ophthalmic data entry errors are automatically repaired by using a data type and business rule-based automatic repair algorithm.

[0024] The system log management submodule uses the log recording framework: Log4j, Slf4j. The log output format and storage path are configured through Log4j to record the operation log of the software system in detail, covering every detail of ophthalmic data collection, storage, transmission, security management and monitoring operations, so as to conduct system troubleshooting and statistical analysis, such as the source address of data collection (accurate to the IP address of the ophthalmic examination equipment, the directory of the server where the medical record documents are located), the collection time (accurate to milliseconds), and the amount of collected data (in bytes); the storage location of ophthalmic data storage (such as which table space the data is stored in the relational database, and which path the file is stored in the distributed file system), storage time, and storage data format; the source platform of data transmission (such as the source server host name or the examination equipment Number), target platform (such as the target mobile terminal device number or the name of the receiving server), transmission protocol (details include UDP protocol and application layer custom protocol details), transmission data volume; encryption operations for ophthalmic data security management (involved ophthalmic data, encryption algorithm and key information), access permission settings (targeted user roles and specific permissions), backup and recovery records (when full backup was performed, the content of incremental backup, and the time and results of recovery operations); data monitoring traffic anomaly information (time when traffic exceeds the threshold, traffic peak, etc.), quality problem data (details of problematic ophthalmic data, problem type and processing results), so as to accurately locate the root cause of the problem when troubleshooting the system and comprehensively trace the history of ophthalmic data operations when conducting statistical analysis, providing a basis for data management and optimization. Through the efficient retrieval function of the log management system, the average response time for querying logs related to specific events in massive log data can be controlled within 2 seconds, which improves the utilization efficiency of log data, such as quickly finding detailed information on a data transmission anomaly.

[0025] An ophthalmic multi-source heterogeneous data fusion storage and cross-platform transmission device, comprising:

[0026] An ophthalmic data acquisition device, used to perform the functions of the ophthalmic data acquisition module and acquire data of different structures from a plurality of different ophthalmic data sources;

[0027] An ophthalmic data storage device, used to execute the functions of an ophthalmic data fusion storage module, construct an ophthalmic data storage model, perform data fusion storage, and provide a storage interface;

[0028] An ophthalmic data transmission device, used to execute the functions of the ophthalmic data transmission module and realize seamless ophthalmic data transmission across platforms;

[0029] An ophthalmic data security device, used to execute the functions of the ophthalmic data security management module and ensure the security of ophthalmic data during storage and transmission;

[0030] The ophthalmic data monitoring device is used to execute the functions of the ophthalmic data monitoring and management module to monitor and manage the ophthalmic data processing process.

[0031] A computer device is composed of a memory and a processor. A computer program is stored in the memory, and when the processor runs the computer program, each step included in any of the above-mentioned methods can be implemented one by one, thereby achieving the corresponding ophthalmic data processing function and task execution requirements.

[0032] A computer-readable storage medium has the ability to store computer programs. When the computer program stored on the medium is called and executed by a processor, each step of any of the above-described ophthalmic data processing methods will be executed in sequence to complete the relevant operation process and achieve the expected target effect.

[0033] (III) Beneficial effects

[0034] The present invention provides a method, device, computer equipment and storage medium for fusion storage and cross-platform transmission of multi-source heterogeneous ophthalmic data. It has the following beneficial effects:

[0035] The method, device, computer equipment and storage medium for fusion storage and cross-platform transmission of multi-source heterogeneous ophthalmic data can effectively solve the problem of fusion storage and cross-platform transmission of multi-source heterogeneous ophthalmic data through close collaboration and fine operation among modules, and can play an important role in multiple key fields such as ophthalmic clinical diagnosis, telemedicine collaboration, medical research and innovation, ophthalmic medical teaching and training, and hospital information management; it solves the problems of difficulty in fusion storage and inconvenience in cross-platform transmission of multi-source heterogeneous ophthalmic data in the prior art, and improves the comprehensive utilization value of ophthalmic data. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 A schematic diagram of a multi-source heterogeneous data collection process for ophthalmology in an embodiment;

[0037] Figure 2 It is a schematic diagram of the fusion and storage process of multi-source heterogeneous data in ophthalmology in one embodiment;

[0038] Figure 3 This is a schematic diagram of the ophthalmic multi-source heterogeneous data transmission process in one embodiment;

[0039] Figure 4 It is a schematic diagram of a flow chart of a method for encrypting multi-source heterogeneous data in ophthalmology in one embodiment;

[0040] Figure 5 It is a flowchart of a method for controlling access rights of multi-source heterogeneous ophthalmological data in one embodiment;

[0041] Figure 6 A schematic diagram of a flow chart of a method for backing up and restoring multi-source heterogeneous data in ophthalmology in an embodiment;

[0042] Figure 7 A schematic diagram of a flow chart of a method for monitoring multi-source heterogeneous data flow in ophthalmology in an embodiment;

[0043] Figure 8 A schematic diagram of a method for monitoring quality of multi-source heterogeneous ophthalmic data in one embodiment;

[0044] Fig. 9 A flow chart of a log management method for an ophthalmic multi-source heterogeneous data system in one embodiment;

[0045] Fig.10 It is a structural block diagram of an ophthalmic multi-source heterogeneous data fusion storage and cross-platform transmission device in one embodiment;

[0046] Fig.11 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0047] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] In an ophthalmic data collection embodiment, Figure 1 As shown, a method for collecting multi-source heterogeneous ophthalmic data is provided, which specifically includes:

[0049] After the S11 data acquisition module is started, the data source identification submodule conducts a comprehensive scan of various potential data sources, including database systems (checking multi-dimensional information such as port numbers, connection string features, such as the port and connection features of the ophthalmic patient information database), file systems (analyzing file path formats, file header information, such as relevant information of the file system where ophthalmic medical records are located), network data sources (based on network request URL format, data return format, and protocol features, such as network features when obtaining data from a remote ophthalmic medical data platform), and specific application programming interfaces (parsing interface specifications and identification information, such as specific identification of ophthalmic examination equipment software interfaces). Determine the data source type based on the identification results. If it is a database system, go to step S12; if it is a file system, go to step S13; if it is a network data source, go to step S14; if it is a specific application programming interface, go to step S15;

[0050] S12 Structured database data extraction submodule uses JDBC or ODBC database connection tools to connect to the database. Construct and execute SQL query statements to extract structured data: including patient basic information table, intraocular pressure measurement data table, and vision test result table. After the extraction is completed, go to step S16;

[0051] Specifically, in the hospital network environment, it detected a data source from a MySQL database, and through in-depth analysis of its port number 3306 and database connection string characteristics, accurately determined that it was a MySQL database type. Then, the data extraction submodule used JDBC to connect to the database, carefully constructed and executed the corresponding SQL query statement, and successfully extracted the patient's name, age, medical history information, as well as previous intraocular pressure measurement data, vision test data and other structured database data.

[0052] S13: Extraction of unstructured document data. If it is a Word document, use text reading technology to read the text content, and adopt a multi-threaded parallel reading strategy to increase the reading speed. If it is a PDF document, call the iText parsing library to parse the document, extract text, pictures, and tables, and use the document object model DOM parsing technology to deeply parse the document structure. After the extraction is completed, go to step S16;

[0053] Specifically, on the hospital file server, the data source identification submodule keenly discovered some ophthalmology medical records. By carefully analyzing the file header information, it was accurately determined to be unstructured document data. For the Word format medical record summary documents, the data extraction submodule uses text reading technology and multi-threaded parallel processing to systematically extract text content, such as doctors' diagnosis analysis, treatment recommendations, etc.; and for the PDF format surgical report documents, it uses a special iText library combined with DOM parsing technology for in-depth analysis, and comprehensively extracts the text descriptions, surgical pictures, postoperative recovery status tables and other content information in the documents.

[0054] S14 Multimedia data extraction: For image data, call the OpenCV library to read pixel information, color features, size, etc., and use efficient image pixel processing algorithms to quickly obtain detailed information. For audio and video data, enable the FFmpeg library to extract channel information, sampling rate, frame rate, resolution, encoding format, etc., and improve extraction efficiency through hardware acceleration technology. After the extraction is completed, go to step S16;

[0055] Specifically, from ophthalmic examination equipment (such as fundus cameras, OCT equipment, etc.) or hospital image storage systems, the data source identification submodule quickly determines the existence of some ophthalmic images and video data, such as fundus images, OCT images, FFA videos, UBM videos and other multimedia data sources based on file format characteristics and device identification information. For image data, the data extraction submodule calls the OpenCV library to accurately read the image's pixel information, color features, size and other detailed information, such as the distribution and morphological characteristics of blood vessels in fundus images and the hierarchical structure information of the retina in OCT images; for audio and video data, the FFmpeg library is enabled to efficiently extract key data such as the video's frame rate, resolution, encoding format, and key frame information in the image sequence, such as the key frame image of the contrast agent flowing in the fundus blood vessels in the FFA video, providing comprehensive data support for the diagnosis of ophthalmic diseases.

[0056] S15 specific application interface data extraction, according to the interface specifications and protocols, use the corresponding interface call method to obtain data, and use the cache mechanism and asynchronous request technology to improve data acquisition efficiency. After the extraction is completed, go to step S16;

[0057] Specifically, the data source identification submodule determines that it is an application program interface data source of an ophthalmology examination device or a medical record management system based on the URL path " / api / ophthalmologyData" of the interface and the request header identifier "Ophthalmology-API-Token". The data extraction submodule constructs a request parameter containing a patient number or an examination item number, sends a request to the interface, stores common data through a cache mechanism, and uses asynchronous request technology to process multiple requests concurrently. After receiving the return data (JSON format), the examination details are extracted, such as parameter settings of specific examination equipment, detailed numerical values ​​of examination results, and other information for subsequent processing.

[0058] S16 data is transferred to the fusion storage module, the extracted data is transferred to the data fusion storage module, and a high-speed data transmission channel and data verification mechanism are used to ensure the accuracy and integrity of data transmission, and the process ends.

[0059] In a data fusion storage embodiment, such as Figure 2 As shown, a method for fusion storage of multi-source heterogeneous ophthalmic data is provided, which specifically includes:

[0060] The data collected by S21 is transmitted to the data fusion storage module in a timely manner. The metadata description unit quickly starts working to collect metadata information of the data. The metadata information of various types of data is collected through the metadata description unit, and stored and managed according to the "Metadata" table structure definition. Regardless of whether the data is a structured patient information table, an unstructured medical record document, or multimedia data, there are corresponding metadata to describe its source (such as examination equipment from an ophthalmology clinic, a doctor's medical record, etc.), type (structured, unstructured, multimedia), format (such as database table structure, Word document format, JPEG image format, etc.), creation time and association relationship (such as the patient's fundus image and the corresponding medical record document association, OCT image and intraocular pressure examination data association, etc.) and other key attributes. If it is structured data, go to S22; if it is unstructured data, go to S23; if it is multimedia data, go to S24;

[0061] Specifically, taking the patient information table as an example, its data source is marked as a specific patient information table in the MySQL database, the data type is clearly structured, the format is recorded in detail as the database table structure, the creation time is accurately recorded as the creation time of the database record, and the association relationship is sorted out through the intelligent association analysis algorithm to find out the possible associations with other eye examination related tables, such as association with the intraocular pressure examination data table through the patient number, etc. These metadata information are strictly defined in accordance with the "Metadata" table structure and inserted into the metadata description layer. The metadata storage uses a compression algorithm to save storage space.

[0062] The data storage unit classifies and stores the data according to the data type.

[0063] S22: For structured data, such as patient information table, intraocular pressure measurement data table, vision test result table, etc., use the InnoDB engine to store them in the relational database storage area, strictly follow the database design specifications, and ensure the integrity and consistency of the data through B+ tree index combined with index coverage. After the step, go to S25;

[0064] S23 Unstructured document data, such as medical record summary documents in Word format and surgical report documents in PDF format, are stored in a specific directory of the Ceph distributed file system in an orderly manner. According to the pre-designed directory structure, the storage is segmented according to factors such as department and medical record type. At the same time, the storage path information of the document is recorded in detail in the metadata description layer, and the high availability and load balancing characteristics of the distributed file system are used to ensure reliable data storage and fast access. The step ends and goes to S25;

[0065] S24 multimedia data, AMD-related fundus images, DR diabetic retinopathy fluorescence angiography videos, FFA fundus fluorescence angiography image sequences, FP fundus photographs, HR hypertensive retinopathy imaging data, MRI magnetic resonance imaging data, OCT optical coherence tomography images, RD retinal detachment image records, RVD retinal vein occlusion images, UBM ultrasound biomicroscopy images and videos, FFA videos, UBM images and videos, based on MinIO object storage, are stored in the multimedia storage area in the form of objects. Each object is assigned a unique object key, and the mapping relationship between the object key and the metadata information such as the resolution, duration, and encoding format of the multimedia data is recorded in the metadata description layer. The multimedia data index construction technology based on the combination of hash index and inverted index is used for query. The step ends and goes to S25;

[0066] During the storage process of S25, the index construction unit actively constructs data indexes. For example, it constructs an index for the patient information table based on keywords such as patient name and medical record number, and adopts a dynamic index update strategy; it constructs an unstructured document index based on the keywords of the document, and uses natural language processing technology to extract keywords, such as extracting disease names, treatment methods and other keywords from medical record documents; it constructs a multimedia data index based on the color features of the image, the retinal layering features of the OCT image, etc., and uses deep learning algorithms to extract feature vectors to construct feature indexes, which improves data query efficiency and helps doctors quickly retrieve and analyze patients' ophthalmic data.

[0067] In one data transmission embodiment, Figure 3 As shown, a method for transmitting multi-source heterogeneous data in ophthalmology is provided, which specifically includes:

[0068] S31 Data transmission preparation stage: when there is a need for data transmission, the platform detection submodule of the data transmission module uses a combination of system detection technology and network scanning technology to comprehensively detect the data transmission source platform (including operating system, device type, such as Windows system of hospital server, iOS or Android system of doctor's mobile terminal, etc.) and target platform (including operating system, device type, network environment information, such as operating system, device model, current network bandwidth, delay, etc. of doctor's tablet or smart phone), and then the transmission protocol adapter submodule selects the appropriate transmission protocol according to the detection results and optimizes the application layer custom protocol;

[0069] Specifically, when data needs to be transmitted to the doctor's mobile terminal, for example, from the hospital server to the doctor's tablet computer. The platform detection submodule of the data transmission module quickly detects that the source platform is the hospital server (operating system is Windows, device type is server), and the target platform is the doctor's tablet computer (operating system is iOS, device type is tablet computer, network bandwidth is 3Mbps, and delay is 30ms). Based on this information, the transmission protocol adaptation submodule decisively selects the lightweight UDP protocol and performs careful custom protocol optimization at the application layer to balance transmission efficiency and data reliability to meet the doctor's needs to quickly view patients' ophthalmic data in mobile scenarios.

[0070] In the data format conversion stage S32, the data format conversion submodule builds a dynamic format recognition and conversion mechanism for different ophthalmic data types. When the system recognizes ophthalmic image data, it uses a deep learning-based image recognition algorithm to detect the image content and determine the image compression and format conversion strategy. If the transmission is interrupted, it enters the breakpoint resume process S33;

[0071] Specifically, in the data collection stage, we first need to use intelligent data source identification technology to determine the source type of each data and classify it. The source types include structured, unstructured, image, and video. This step includes:

[0072] Step 1: Acquire data from multiple ophthalmic data sources, including databases (MySQL, Oracle), file systems (PDF, Word documents), image data sources (fundus images, OCT images), and video data sources (FFA videos, OCT videos). The process includes: determining the data source through port scanning, file header analysis, and interface protocol parsing. For database data, it is determined to be structured data (MySQL, Oracle) based on the port number and connection string features. For file system data, it is determined to be unstructured document data through file header information and file type identification. For image and video data, an image recognition algorithm (deep learning CNN) is used to analyze the file type and output the data type, including structured, unstructured, image, and video. The identified data is classified according to the type and enters the subsequent processing flow, including image processing, video compression, and document parsing.

[0073] Step 2: For different types of data, format conversion and adaptation are performed according to the requirements of the target platform. This step is processed according to the specific content of the data and the requirements of the target device (display device, transmission environment). For image data processing, high-resolution ophthalmic images (fundus images, OCT images) are input. The process includes: analyzing the image content, using convolutional neural network (CNN) to analyze the image content, and identifying the lesion area in the image (retina, vascular distribution); judging whether compression or resolution adjustment is required based on the image content: for background images or images without lesions, a high compression ratio (WebP format) is used to save storage and transmission bandwidth. For images with key lesion areas, more details are retained, a lower compression ratio is used, and the JPEG format is used. The image format adapted to the target device (WebP, JPEG) is output;

[0074] For video data processing: input ophthalmic videos (FFA videos, OCT videos), analyze the video content: extract the video frame rate, resolution, and encoding format through the FFmpeg library, adjust the video resolution and bit rate according to the network bandwidth and the processing capacity of the device, and select a low-resolution, high-compression video format (H.264 to H.265) for low-bandwidth or small-screen devices. For environments with sufficient bandwidth and high quality requirements, retain high-resolution video content, select higher-quality encoding (MP4), and output a video format suitable for the device and network conditions (low-bitrate H.264 video, low-resolution MP4 video);

[0075] For document data processing: input ophthalmology medical record documents (PDF, Word documents), use PDF parsing library (iText) or document parsing technology (regular expression, DOM parsing) to extract key information from the documents, select the appropriate file format conversion according to the display requirements of the target platform (mobile device or PC), convert the documents from PDF or Word format to structured format (JSON or XML) for machine reading and fast processing, and output standardized structured document format (JSON, XML);

[0076] Step 3: After format recognition and conversion, the system dynamically adjusts the data compression method and format according to the target device, network conditions and data characteristics to ensure data transmission efficiency and user experience;

[0077] For image compression and adaptation: Input: image data (fundus image, OCT image). If the target device supports high-resolution images, the system will store and transmit the image at a low compression rate (JPEG or PNG format). If the network bandwidth is low or the target device is a mobile device, the system will compress the image and select a suitable compression format (WebP format) to reduce the amount of transmitted data. Output: compressed image (WebP format) or original image (JPEG format).

[0078] For video compression and adaptation: Input: video data (OCT video, FFA video), select the appropriate encoding format (H.265, H.264) according to the screen resolution and processing power of the target device, make adaptive adjustments to the video, compress the video during transmission (reduce the bit rate, adjust the resolution), output: optimized low-bit rate video format (H.264 or H.265 video);

[0079] For document compression and adaptation: Input: unstructured documents (PDF, Word documents), perform structured conversion on the documents, extract the document content and convert it into a suitable file format (JSON, XML), choose whether to compress according to the complexity of the document content, retain key data, reduce irrelevant content, output: structured document data (JSON, XML format);

[0080] Step 4: In order to ensure the real-time and stability of the format conversion process, the system introduces a real-time monitoring and feedback mechanism to cope with network fluctuations and bandwidth limitations, monitor data traffic and network, input transmission requests for images, videos, and documents, and monitor network bandwidth, latency, and data traffic in real time. When the system detects that the network bandwidth is insufficient or the latency is too high, it automatically adjusts the transmission strategy: compress images more highly, use a format that is more suitable for low-bandwidth environments (WebP), reduce the resolution or bit rate of videos, use low-resolution video stream transmission, and output optimized image, video, and document data transmission streams;

[0081] Step 5: Format adaptation and transmission: transfer the converted and optimized data to the target device or platform, input the converted and optimized data (images, videos, documents), automatically select the appropriate transmission protocol (HTTP, FTP, UDP) according to the needs of the target platform (mobile device, PC, hospital server), ensure that the data can be seamlessly adapted according to the display characteristics of the device and network conditions, and realize data breakpoint continuation. When the transmission process is interrupted, the transmission can be resumed from the interruption point and the data successfully transmitted to the target device or platform can be output;

[0082] The implementation process of dynamic format recognition and conversion includes automatic recognition of data sources, intelligent format conversion, data compression and adaptation, real-time monitoring and feedback mechanisms. Through these steps, ophthalmic data can be seamlessly, efficiently and accurately transmitted and stored on different devices and in different network environments, ultimately achieving the goal of optimizing data transmission efficiency and improving user experience.

[0083] In the S33 breakpoint resume stage, if the transmission is interrupted, the breakpoint resume unit records the breakpoint position and continues to transmit data from the breakpoint according to the record after the network is restored, ensuring that the data is fully transmitted to the target platform.

[0084] Specifically, if the doctor's tablet computer suddenly loses connection with the Internet during the transmission process, the breakpoint resume unit quickly records the breakpoint position of the transmission, such as accurately recording the number of the transmitted data block, the file pointer position and other information. When the tablet computer is reconnected to the Internet, the breakpoint resume unit continues to transmit data from the breakpoint based on the recorded breakpoint position, ensuring that the data is completely transmitted to the doctor's tablet computer, so that the doctor can successfully obtain the patient's complete ophthalmic data without affecting the continuity of the diagnosis work.

[0085] In one data security management embodiment, Figure 4 , provides a data encryption method for secure management of multi-source heterogeneous ophthalmic data, specifically including:

[0086] The S41 data encryption submodule identifies sensitive data and uses data classification and sensitivity tagging technology to quickly locate sensitive data, such as patients' private information (including ID number, home address, contact information, etc.), research data on rare eye diseases, and high-precision eye examination image data;

[0087] For sensitive data, S42 first uses the AES symmetric encryption algorithm to quickly encrypt the data content;

[0088] S43 uses the RSA asymmetric encryption algorithm to encrypt and transmit AES encryption keys to ensure data confidentiality, and adopts an optimized key management strategy to ensure secure key distribution and storage.

[0089] Specifically, the data encryption submodule encrypts sensitive ophthalmic data to ensure data security. For example, for patients' high-definition fundus image data, the AES symmetric encryption algorithm is first used to quickly encrypt the image data content, and then the RSA asymmetric encryption algorithm is used to encrypt the AES encryption key for transmission to ensure the confidentiality of the data during storage and transmission. During the encryption process, the key management server is used to centrally generate and distribute keys, and the keys are regularly updated and backed up, stored, and distributed.

[0090] In one data security management embodiment, Figure 5, provides an access rights control method for the security management of multi-source heterogeneous ophthalmological data, specifically including:

[0091] S51 Access control submodule determines different roles based on the role access control RBAC model. If it is an ophthalmologist role, go to S52; if it is a general ophthalmologist, go to S53; if it is a nurse role, go to S54; if it is a researcher role, go to S55; if it is a patient role, go to S56;

[0092] S52 grants ophthalmologists full access rights (read, write, modify and delete) to all patient data. Through the audit function of the rights management system, detailed records of permission operations are kept to comprehensively control the diagnosis of difficult diseases and the formulation of treatment plans.

[0093] S53 grants the general ophthalmologist role the permission to read and modify some of the data of the patients he is responsible for (such as basic information, routine examination data, preliminary diagnosis results, etc.). It adopts a permission refinement strategy to allocate precise access scope according to the scope of patients the doctor is responsible for, thus ensuring the privacy and security of patient data;

[0094] S54 sets the nurse role to only allow the reading permission of viewing the patient's basic information (such as name, gender, age, etc.) and nursing-related data (such as nursing records, medication status, etc.) so as to carry out daily nursing work;

[0095] S55 is a role setting for researchers, which allows them to read and analyze anonymized ophthalmic research data, but does not allow them to modify or delete data. This prevents data leakage and illegal access, and provides researchers with a secure data access environment through data desensitization technology, thus promoting the progress of ophthalmic medical research.

[0096] S56 grants the patient role the permission to only view his or her own basic examination results (such as vision, intraocular pressure, etc.) and diagnostic information, so that he or she can understand his or her own condition.

[0097] In one data security management embodiment, Figure 6 , provides a data backup and recovery method for secure management of multi-source heterogeneous ophthalmic data, specifically including:

[0098] The S61 data backup and recovery submodule sets a full backup cycle based on the importance and update frequency of the data. For example, it backs up all newly generated ophthalmic examination data every morning and uses an efficient backup task scheduling algorithm to ensure that the backup task is executed on time.

[0099] S62 When data is updated, such as changes in the patient's intraocular pressure test data, the changed data blocks are recorded in time for incremental backup, and data change monitoring technology is used to capture data updates in real time;

[0100] S63 determines the storage location based on the importance of the data. Important data (such as the patient's original examination images, long-term medical records, etc.) are backed up to a remote disaster recovery center. Dedicated line transmission and encrypted storage technology are used to ensure the security and integrity of the remote backup data. Ordinary business data (such as temporary examination record summaries, etc.) are backed up locally.

[0101] When data is lost or damaged, S64 uses backup data and data log-based recovery technology to restore data to a specified point in time to ensure data availability and integrity.

[0102] In one data monitoring and management embodiment, Figure 7 , provides a method for data flow monitoring of multi-source heterogeneous data monitoring and management in ophthalmology, specifically including:

[0103] The S71 data flow monitoring submodule is started, using the flow monitoring tool based on the SNMP protocol and the sliding window algorithm to count the flow data, and using high-performance flow monitoring equipment and real-time data processing algorithms to ensure accurate collection and fast calculation of flow data;

[0104] S72 monitors the flow of data in real time during storage and transmission, and counts the inflow and outflow of data as well as the data transmission rate;

[0105] S73 compares the real-time flow data with a preset flow threshold model (built based on historical flow data and ophthalmic data processing system performance indicators), and the flow threshold model uses an intelligent dynamic adjustment strategy to optimize the threshold in real time according to the system operation status;

[0106] If the data flow exceeds the threshold, S74 will promptly issue an alarm and initiate a troubleshooting process to locate and resolve the abnormal flow problem, such as checking the network bandwidth usage and determining whether there is malicious data transmission or network failure. During the flow monitoring process, a real-time flow data analysis algorithm is used, which can issue an alarm within 1 second after the flow abnormality occurs, and the false alarm rate is controlled below 2%.

[0107] Specifically, Figure 8 As shown, a method for monitoring the quality of multi-source heterogeneous ophthalmic data is provided, which specifically includes:

[0108] The S81 data quality monitoring submodule is loaded with a complete data verification rule library and intelligent data cleaning and repair algorithms. The data verification rule library adopts a modular design and can flexibly expand the verification rules;

[0109] S82 comprehensively checks the integrity of the stored data, including whether the required fields in various data forms (such as patient information forms) are complete and whether there are any missing ophthalmic examination data records;

[0110] S83 verifies data accuracy, and checks the numerical correctness and data format standardization of key business data (such as whether intraocular pressure measurement data is within a reasonable range, whether the diagnostic information in the medical record is accurate and standardized, etc.);

[0111] S84 performs data consistency comparison to check whether the same data such as vision data and intraocular pressure data of the same patient generated by different examination equipment are consistent;

[0112] S85 When it is found that the data quality does not meet the standard, the corresponding processing flow is automatically triggered according to the type of data problem, such as starting the data cleaning process for duplicate records, using efficient duplicate detection algorithms to remove duplicate data; repairing the input error data, using data association relationships and ophthalmic expertise and business rules for intelligent repair; deleting serious errors or expired data to ensure that the data quality meets the needs of ophthalmic medical and scientific research. In the data quality monitoring test, for large-scale ophthalmic data samples, the accuracy of data integrity verification can reach more than 99%, the accuracy of accuracy verification can reach more than 98%, and the accuracy of consistency verification can reach more than 97%.

[0113] Specifically, Fig. 9 , provides a log management method for ophthalmology multi-source heterogeneous data system, specifically including:

[0114] The S91 system log management submodule initializes the efficient log recording framework Log4j and uses log classification and classification storage strategies to optimize log management;

[0115] S92 records detailed operation log information at each key link of data processing, including the source address of the data collection link (accurate to the IP address of the ophthalmic examination equipment and the directory of the server where the medical record documents are located), the collection time (accurate to milliseconds), and the amount of collected data (in bytes);

[0116] S93 records the storage location (table space of relational database, path of distributed file system), storage time, and storage data format of the data storage link;

[0117] S94 records the source platform (source server host name or inspection device number), target platform (device number of the target mobile device or receiving server name), transmission protocol (details including UDP protocol and application layer custom protocol details), and transmission data volume of the data transmission link;

[0118] S95 records the encryption operations in the data security management process (involved ophthalmic data, encryption algorithms and key information), access permission settings (targeted user roles and specific permissions), and backup and recovery records (time, content and recovery operation details of full and incremental backups);

[0119] S96 records abnormal flow information (time when flow exceeds threshold, flow peak, etc.) and quality problem data (details of problematic ophthalmic data, problem type, and treatment results) in the data monitoring link, so that the data operation history can be fully traced during subsequent system troubleshooting and audit analysis. Through the efficient retrieval function of the log management system, the average response time for querying logs related to specific events in massive log data can be controlled within 2 seconds.

[0120] In one embodiment, Fig.10 As shown, an ophthalmic multi-source heterogeneous data fusion storage and cross-platform transmission device is provided, including: a data acquisition module, a data fusion storage module, a data transmission module, a data security management module, and a data monitoring and management module, wherein:

[0121] The data acquisition module is used to collect data of different structures from multiple different ophthalmic data sources, including obtaining data from ophthalmic patient information databases, ophthalmic medical record document systems, various ophthalmic examination equipment: fundus cameras, OCT equipment, FFA equipment, UBM equipment, and related network data sources and specific application program interfaces;

[0122] A data fusion storage module is used to fuse and store the collected ophthalmic multi-source heterogeneous data according to the data storage model, and provide a data storage interface so that an external system can access the stored ophthalmic data, so as to facilitate data management and call;

[0123] The data transmission module is used to achieve seamless data transmission between different operating systems: Windows system for hospital servers, iOS or Android system for doctors' mobile terminals, and different types of devices: large ophthalmic examination equipment, desktop computers, laptops, tablet computers, and smart phones, meeting the data interaction needs between diverse devices and systems in ophthalmic medical scenarios;

[0124] Data security management module, used to encrypt ophthalmic data during storage and transmission, control access rights, and perform data backup and recovery to ensure data security and integrity;

[0125] The data monitoring and management module is used to monitor the data in real time during storage and transmission, monitor the quality of stored ophthalmic data, and record the operation log of the software system to ensure the stability and reliability of the data processing process.

[0126] In terms of the relevant settings of the ophthalmic multi-source heterogeneous data fusion storage and cross-platform transmission device, the limitations given in the previous article for the multi-source heterogeneous data fusion storage and cross-platform transmission method can be referred to, and no further details will be given here. Each module inside the above-mentioned multi-source heterogeneous fusion storage and transmission device can be implemented through software, hardware, or a combination of the two. These modules can be directly embedded in the processor of the computer device in the form of hardware, or they can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operation process corresponding to each module when needed.

[0127] In one embodiment, a computer device is provided. The device may be a terminal type. The internal structure diagram thereof can be referred to as Fig.11 The computer device connects the processor, memory, network interface, display screen and input device to each other with the help of the system bus. Among them, the processor is responsible for providing computing power and control functions. The memory of the computer device includes two parts: non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and related computer programs, while the internal memory creates a suitable environment for the normal operation of the operating system and computer programs in the non-volatile storage medium. The network interface of the computer device enables the device to connect with other external ophthalmology-related terminals through the network and conduct communication interaction. When the computer program is executed by the processor, a specific method of fusion storage and cross-platform transmission of multi-source heterogeneous data in ophthalmology can be achieved. The display screen of the computer device is a liquid crystal display screen or an electronic ink display screen, and its input device can be a touch layer covering the display screen, or a key, trackball or touchpad set on the shell of the computer device, or even an external device such as an external keyboard, touchpad or mouse.

[0128] Professionals and technicians in this field should know that Fig.11 The structure shown in the figure is only a partial structural framework diagram related to the present application scheme, and it does not constitute an absolute limitation on the computer device to which the present application scheme is applied. In fact, the specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different component layout.

[0129] In one embodiment, a computer device is provided, which includes a memory and a processor. The memory stores a computer program. When the processor executes the computer program, each step in the above method embodiment can be implemented one by one, thereby achieving efficient processing and management of multi-source heterogeneous ophthalmic data.

[0130] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. Once the computer program is executed by a processor, each step of the above method embodiment is executed in sequence, providing a solid software operation foundation for ophthalmic data processing.

[0131] Those of ordinary skill in the art will understand that if all or part of the operation flow in the above-mentioned embodiment method is to be implemented, instructions can be issued to the relevant hardware to complete the task by means of a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. During the execution of the computer program, the process content contained in the above-mentioned various method embodiments will be involved. It should be noted that any reference to memory, storage, database or other media in the various embodiments provided in this application can cover non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory, etc. For example, RAM exists in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM) and memory bus dynamic RAM (RDRAM).

[0132] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "including a..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0133] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for fusion storage and cross-platform transmission of multi-source heterogeneous ophthalmic data, characterized in that: include: A data collection module, used for collecting ophthalmic multi-source heterogeneous data from multiple different sources, wherein the ophthalmic multi-source heterogeneous data includes ophthalmic structured database data, ophthalmic unstructured document data and ophthalmic multimedia data; A data fusion storage module is used to construct a data storage model for ophthalmic data, fuse and store the collected ophthalmic multi-source heterogeneous data according to the data storage model, and provide a data storage interface; The data transmission module is used to achieve seamless data transmission across different operating systems and different device types. During the data transmission process, ophthalmic data format conversion and adaptation are automatically performed according to the characteristics of the target platform; The data storage model in the data fusion storage module includes: The metadata description layer is used to store metadata information of multi-source heterogeneous ophthalmology data, including the source of ophthalmology data, data type, data format, data creation time and data association relationship. The metadata information is stored in the "OphthalmologyMetadata" table structure in the form of key-value pairs; The data storage layer classifies and stores ophthalmic data in different data storage areas according to data types. For ophthalmic structured data, a relational database storage method is used, and the InnoDB engine is used to store it in the relational database storage area, and the ophthalmic database design specifications are strictly followed. For ophthalmic unstructured document data, a file storage system is used for storage, and it is stored in a specific ophthalmic directory of the Ceph distributed file system in an orderly manner, and is subdivided and stored according to the preset ophthalmic directory structure rules. For ophthalmic multimedia data, an ophthalmic multimedia repository is used for storage, and it is stored in the ophthalmic multimedia storage area in the form of objects based on MinIO object storage. Each object is assigned a unique object key, and a mapping relationship is established between the data storage layer and the metadata description layer, and the mapping of metadata and stored data is achieved through a data structure based on a hash table or index tree; The data transmission module comprises: The platform detection submodule is used to detect the operating system type, device type and network environment information of the source platform and target platform of data transmission; The transmission protocol adapter submodule selects the appropriate transmission protocol for ophthalmic data transmission based on the information detected by the platform detection submodule. A lightweight transmission protocol is used for low-bandwidth network environments, and a reliable transmission protocol is used for high-bandwidth environments with high data integrity requirements. The data format conversion submodule builds a dynamic format recognition and conversion mechanism for different ophthalmic data types. When the system recognizes ophthalmic image data, it uses a deep learning-based image recognition algorithm to detect the image content and determine the image compression and format conversion strategy. The process of building a dynamic format recognition and conversion mechanism includes: source identification and classification of input ophthalmic data through the data acquisition module; after the data is classified, identify the characteristics of each data format, select the conversion method and format for the type of data, compress and convert according to the type and target device requirements, design a real-time monitoring and feedback mechanism, take image, video, and document transmission requests as input, monitor network bandwidth, delay and data flow in real time, and automatically adjust the transmission strategy when the system detects that the network bandwidth is insufficient or the delay is too high, output the optimized image, video, and document data transmission stream, and transmit the converted and optimized data to the target device or platform.

2. The method for fusion storage and cross-platform transmission of multi-source heterogeneous ophthalmic data according to claim 1, characterized in that: The data acquisition module comprises: A data source identification submodule, which is used to automatically identify the source type of ophthalmic data, including ophthalmic database systems, ophthalmic file systems, ophthalmic network data sources, and specific ophthalmic application program interfaces; The data extraction submodule extracts ophthalmic data from the data source using corresponding extraction techniques according to the source type determined by the data source identification submodule. For ophthalmic structured database data, SQL query statements or database connection tools are used for extraction. Specifically, JDBC connection is used for specific ophthalmic databases and accurate SQL query statements are constructed based on the database table structure to extract data. For ophthalmic unstructured document data, document parsing technology is used to extract text content. For ophthalmic PDF documents, iText library is used to parse and extract text, pictures, and tables. For ophthalmic multimedia data, a special multimedia data reading library is used for extraction. For FFA fluorescence angiography videos and UBM ultrasonic biomicroscopy videos, the FFmpeg library is enabled to extract channel information, sampling rate, video frame rate, resolution, and encoding format.

3. The method for fusion storage and cross-platform transmission of multi-source heterogeneous ophthalmic data according to claim 2 further comprises a data security management module, characterized in that: The data encryption submodule is used to encrypt the ophthalmic data during storage and transmission. It uses a combination of symmetric encryption algorithms and asymmetric encryption algorithms to encrypt sensitive ophthalmic data. During the encryption process, a key management server is used to centrally generate and distribute keys, and the keys are regularly updated and backed up, stored, and distributed; The access rights control submodule sets different access rights according to the user role and the sensitivity of ophthalmic data. Only users or systems with corresponding permissions can access specific ophthalmic data. Access rights include read, write, modify and delete permissions. The role-based access control RBAC model uses a specific permission allocation table structure to set and manage permissions. The data backup and recovery submodule regularly backs up the stored ophthalmic data and restores the data based on the backup data when the data is lost or damaged. The data backup strategy is flexibly set according to the importance and update frequency of the ophthalmic data. The database backup tool mysqldump is used to perform a full backup of the core ophthalmic business data every morning. When the data is updated, the changes in the ophthalmic data are recorded through the data log, and incremental backups are performed based on the logs. The storage location is determined based on the importance of the ophthalmic data. Important ophthalmic data are backed up to an off-site disaster recovery center, while ordinary ophthalmic business data are backed up locally.

4. The method for fusion storage and cross-platform transmission of multi-source heterogeneous ophthalmic data according to claim 3 further comprises a data monitoring and management module, characterized in that: The data flow monitoring submodule monitors the flow of ophthalmic data in real time during storage and transmission, including data inflow, outflow and data transmission rate. When the data flow exceeds the preset threshold, an alarm is issued. The network flow monitoring technology and data flow statistical algorithm are used. The preset flow threshold model is based on the threshold curve constructed by the historical ophthalmic flow data and system performance indicators. The data quality monitoring submodule monitors the quality of stored ophthalmic data, checks the integrity, accuracy and consistency of the data, and marks and processes the ophthalmic data that does not meet the quality requirements. The processing methods include ophthalmic data cleaning, data repair or data deletion. Among them, a data verification rule base and a data cleaning and repair algorithm obtained by storing verification rules in the form of a database table are used; The system log management submodule records the operation log of the software system, including detailed information on ophthalmic data collection, storage, transmission, security management and monitoring operations. It uses a log recording framework to record log information, including the source address, collection time, and amount of collected data for ophthalmic data collection; the storage location, storage time, and storage data format of ophthalmic data storage; the source platform, target platform, transmission protocol, and amount of transmitted data for ophthalmic data storage; encryption operations, access permission settings, and backup and recovery records for ophthalmic data security management; and traffic anomaly information and quality problem data for ophthalmic data monitoring.

5. An ophthalmic multi-source heterogeneous data fusion storage and cross-platform transmission device, characterized by: include: The data acquisition module is used to collect data of different structures from multiple different ophthalmic data sources, including obtaining data from ophthalmic patient information databases, ophthalmic medical record document systems, various ophthalmic examination equipment: fundus cameras, OCT equipment, FFA equipment, UBM equipment, and related network data sources and specific application program interfaces; A data fusion storage module is used to fuse and store the collected ophthalmic multi-source heterogeneous data according to the data storage model, and provide a data storage interface for data management and call; Data transmission module, used to achieve seamless data transmission between different operating systems and different types of devices, and to meet the data interaction requirements between diverse devices and systems in ophthalmic medical scenarios; Data security management module, used to encrypt ophthalmic data during storage and transmission, control access rights, and perform data backup and recovery to ensure data security and integrity; The data monitoring and management module is used to monitor the data in real time during storage and transmission, monitor the quality of stored ophthalmic data, and record the operation log of the software system to ensure the stability and reliability of the data processing process.

6. A computer device, characterized in that: Its structure covers a memory, a processor, and a computer program stored in the memory and capable of running on the processor. When the processor executes the computer program, it can achieve the functions of the ophthalmic multi-source heterogeneous data fusion storage and cross-platform transmission method as described in any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that: This storage medium stores a computer program, which, when executed by a processor, can prompt the processor to execute the functions of the method for fusion storage and cross-platform transmission of ophthalmic multi-source heterogeneous data as described in any one of claims 1 to 4.

Images