AI-based automatic generation and correction system for visual optical medical record

Abstract

The invention discloses an AI-based eye vision optical medical record automatic generation and correction system which comprises a core processing unit, and the output end of the core processing unit is in electric signal connection with a data acquisition module, a semantic analysis module, a dynamic correction module, a data synchronization module and a user interaction interface. And the output ends and the input ends of the data acquisition module, the semantic analysis module, the dynamic correction module, the data synchronization module and the user interaction interface are in bidirectional electric signal connection with the input end and the output end of the core processing unit. According to the AI-based visual optical medical record automatic generation and correction system, a data acquisition module is arranged, specific structured data in the visual optical field is connected through a structured data interface, and meanwhile, free text information input by a doctor is obtained through an unstructured data interface; and the real-time data flow interface is combined to realize real-time acquisition of dynamic data of the patient, so that comprehensive integration of multi-source heterogeneous data is realized, and a complete data basis is provided for subsequent medical record generation.

Description

Technical Field

[0001] This invention relates to the field of medical information technology, specifically to an AI-based automatic generation and correction system for optometry medical records. Background Technology

[0002] With the widespread application of artificial intelligence technology in the medical field, the technology for automatic generation and correction of electronic medical records has developed rapidly. This is particularly true in the field of optometry, where doctors need to process large amounts of patient data and generate detailed medical records, making automation and intelligence essential. However, existing electronic medical record generation technologies still have shortcomings when addressing the specific needs of optometry, especially in terms of accuracy, real-time performance, and personalized correction of generated medical records.

[0003] A method and system for generating electronic medical records, with publication number CN104485105B and publication date April 13, 2018, describes a method and system for generating electronic medical records. This patent utilizes voice acquisition and recognition technology to generate electronic medical records, processing sound feature files using acoustic models, N-gram speech models, and neural network language models to achieve efficient record generation. However, this technical solution relies heavily on voice input and does not fully consider the automated integration capabilities of structured data specific to the field of optometry (such as visual acuity test results and corneal topography analysis). Furthermore, the solution lacks intelligent correction functions for the generated medical record content, potentially leading to inaccurate or incomplete information, especially when highly specialized medical terminology is required.

[0004] A method and apparatus for real-time medical record generation, with publication number CN116913450B, was published on December 19, 2023. This patent achieves real-time generation of complete medical records by training a large-scale language model and combining doctor-patient dialogue text with annotated medical record data. However, this technical solution has limitations when applied to the field of ophthalmology: firstly, the model's training data is mainly based on general medical scenarios and lacks in-depth optimization for ophthalmology, which may prevent accurate processing of ophthalmology-related professional terminology and complex data; secondly, although this solution can generate medical records in real time, it lacks a dynamic correction mechanism for the generated content, making it difficult to update and improve the medical records based on subsequent supplementary examination data, thus affecting the accuracy and practicality of the medical records.

[0005] The aforementioned problems indicate that existing electronic medical record generation technologies still suffer from insufficient data integration capabilities, limited specialization, and a lack of medical record correction functions when addressing the specialized needs of the optometry field. Therefore, this invention provides an AI-based automatic generation and correction system for optometry medical records. It aims to improve the accuracy and practicality of medical record generation by integrating optometry-specific data, optimizing natural language processing models, and introducing a dynamic correction mechanism, thereby meeting the demand for efficient and intelligent medical record management in optometry diagnosis and treatment. Summary of the Invention

[0006] The purpose of this invention is to provide an AI-based automatic generation and correction system for optometric medical records to solve the problems mentioned in the background.

[0007] To achieve the above objectives, the present invention provides the following technical solution: an AI-based automatic generation and correction system for optometry medical records, the system comprising a core processing unit, the output of the core processing unit being electrically connected to a data acquisition module, a semantic parsing module, a dynamic correction module, a data synchronization module, and a user interface, the output and input ends of the data acquisition module, the semantic parsing module, the dynamic correction module, the data synchronization module, and the user interface being bidirectionally electrically connected to the input and output ends of the core processing unit.

[0008] Preferably, the output terminal of the data acquisition module is electrically connected to a structured data interface, an unstructured data interface, and a real-time data stream interface. The input and output terminals of the structured data interface, the unstructured data interface, and the real-time data stream interface are bidirectionally electrically connected to the output and input terminals of the data acquisition module. The output terminal of the structured data interface is electrically connected to a vision testing device interface, a corneal topography analysis interface, and an intraocular pressure measurement device interface. The output and input terminals of the vision testing device interface, the corneal topography analysis interface, and the intraocular pressure measurement device interface are bidirectionally electrically connected to the input and output terminals of the structured data interface.

[0009] Preferably, the output terminal of the semantic parsing module is electrically connected to a medical terminology mapping unit, a context understanding unit, and a syntactic analysis unit. The input and output terminals of the medical terminology mapping unit, the context understanding unit, and the syntactic analysis unit are bidirectionally electrically connected to the output and input terminals of the semantic parsing module. The output terminal of the medical terminology mapping unit is electrically connected to a terminology database interface and a rule engine interface. The input and output terminals of the terminology database interface and the rule engine interface are bidirectionally electrically connected to the output and input terminals of the medical terminology mapping unit.

[0010] Preferably, the output terminal of the dynamic correction module is electrically connected to an incremental learning unit, a conflict detection unit, and a correction strategy unit. The input and output terminals of the incremental learning unit, the conflict detection unit, and the correction strategy unit are bidirectionally electrically connected to the output and input terminals of the dynamic correction module. The output terminal of the incremental learning unit is electrically connected to a model update interface and a data feedback interface. The input and output terminals of the model update interface and the data feedback interface are bidirectionally electrically connected to the output and input terminals of the incremental learning unit.

[0011] Preferably, the output terminal of the data synchronization module is electrically connected to a cloud storage interface, a local storage interface, and a data backup interface. The input and output terminals of the cloud storage interface, the local storage interface, and the data backup interface are bidirectionally electrically connected to the output and input terminals of the data synchronization module. The output terminal of the cloud storage interface is electrically connected to a distributed database interface and an encrypted transmission interface. The input and output terminals of the distributed database interface and the encrypted transmission interface are bidirectionally electrically connected to the output and input terminals of the cloud storage interface.

[0012] Preferably, the output terminal of the user interface is electrically connected to a visualization display unit, an operation control unit, and a feedback collection unit. The input and output terminals of the visualization display unit, the operation control unit, and the feedback collection unit are bidirectionally electrically connected to the output and input terminals of the user interface. The output terminal of the visualization display unit is electrically connected to a chart generation interface and a text rendering interface. The input and output terminals of the chart generation interface and the text rendering interface are bidirectionally electrically connected to the output and input terminals of the visualization display unit.

[0013] Preferably, the structured data interface is configured to receive standardized data from the vision testing device interface, the corneal topography analysis interface, and the intraocular pressure measurement device interface.

[0014] Preferably, the medical terminology mapping unit calls a preset ophthalmology terminology database through a terminology database interface to achieve data conversion.

[0015] Preferably, the incremental learning unit receives the latest training data through the model update interface and updates its internal algorithm online in conjunction with the information provided by the data feedback interface.

[0016] Preferably, the cloud storage interface stores data to the cloud server through a distributed database interface and achieves secure data transmission through an encrypted transmission interface.

[0017] Compared with existing technologies, the beneficial effects of this invention are as follows: This AI-based automatic generation and correction system for optometry medical records, by setting up a data acquisition module, connects to structured data unique to the optometry field through a structured data interface, while simultaneously acquiring free text information input by doctors through an unstructured data interface, and combining it with a real-time data stream interface to achieve real-time acquisition of dynamic patient data, realizes comprehensive integration of multi-source heterogeneous data, and provides a complete data foundation for subsequent medical record generation.

[0018] By setting up a semantic parsing module, the collected data is mapped to a standardized medical terminology system via a medical terminology mapping unit. Combined with a context understanding unit, the medical record content is semantically parsed. At the same time, a syntactic analysis unit is used to optimize and adjust the sentence structure, ensuring the professionalism and accuracy of the generated medical records. It performs particularly well when dealing with complex terms unique to the field of optometry.

[0019] By setting up a dynamic correction module, the incremental learning unit updates the model online based on newly collected data. Combined with the conflict detection unit, logical contradictions in the generated content are identified, and a specific correction plan is formulated through the correction strategy unit. This enables dynamic adjustment of the generated medical record content, ensuring that it maintains accuracy and completeness based on subsequent supplementary data.

[0020] By setting up a data synchronization module, medical record data can be stored and shared in the cloud via a cloud storage interface. Combined with a local storage interface, offline access capability is provided. Furthermore, a data backup interface ensures the security and reliability of the data, thereby improving the management efficiency of medical record data while guaranteeing data security.

[0021] By setting up a user interface, the system presents medical records in a visual format using charts and text through a visualization unit. Combined with an operation control unit, it provides a convenient entry point for operation. Furthermore, by collecting user feedback through a feedback collection unit, the system's usability is improved, and data support is provided for subsequent optimization. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the system architecture. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] Please see Figure 1This invention provides a technical solution: an AI-based automatic generation and correction system for optometric medical records. The system includes a core processing unit and, connected to it via bidirectional electrical signals, a data acquisition module, a semantic parsing module, a dynamic correction module, a data synchronization module, and a user interface. The data acquisition module connects to external devices such as a vision testing device interface, a corneal topography analysis interface, and an intraocular pressure measurement device interface through structured data interfaces, unstructured data interfaces, and real-time data stream interfaces, respectively. The semantic parsing module internally includes a medical terminology mapping unit, a context understanding unit, and a syntactic analysis unit, wherein the medical terminology mapping unit is further connected to a terminology database interface and a rule engine interface.

[0025] Specifically, the dynamic correction module consists of an incremental learning unit, a conflict detection unit, and a correction strategy unit. The incremental learning unit is also connected to the model update interface and the data feedback interface.

[0026] The data synchronization module implements data management functions through cloud storage interfaces, local storage interfaces, and data backup interfaces. The cloud storage interface is further connected to a distributed database interface and an encrypted transmission interface. The user interface consists of a visualization unit, an operation control unit, and a feedback collection unit. The visualization unit connects to a chart generation interface and a text rendering interface.

[0027] In actual operation, the data acquisition module first obtains standardized examination result data from the vision testing device interface, corneal topography analysis interface, and intraocular pressure measurement device interface through structured data interfaces. For example, when a patient undergoes a vision test, the vision testing device interface transmits the patient's uncorrected visual acuity, corrected visual acuity, and other data in structured form to the data acquisition module. Simultaneously, the unstructured data interface receives free text information input by the doctor, such as the patient's chief complaint or diagnostic opinion, while the real-time data stream interface continuously collects dynamic data of the patient during the diagnosis and treatment process, such as pupil changes or intraocular pressure fluctuations. After preliminary processing, this data is transmitted to the core processing unit.

[0028] The core processing unit transmits the received data to the semantic parsing module for in-depth processing. The medical terminology mapping unit uses a terminology database interface to call a pre-defined optometry terminology database to convert the raw data into standardized medical terminology. For example, "uncorrected visual acuity 0.8" is converted to the internationally standardized "Snellen visual acuity 20 / 25". The context understanding unit combines the terminology mapping results to perform semantic-level parsing of the data, ensuring the accuracy of the logical relationships between different data items. The syntactic analysis unit then optimizes the language structure of the final generated text to conform to professional medical record writing standards. The data processed by the semantic parsing module is then returned to the core processing unit.

[0029] The dynamic correction module is responsible for real-time adjustment and optimization of the generated medical record content. The incremental learning unit receives the latest training data through the model update interface and updates the internal algorithm model online by combining the usage feedback information provided by the data feedback interface. When newly acquired data contradicts existing medical record content, the conflict detection unit identifies these logical inconsistencies and formulates specific correction schemes through the correction strategy unit. For example, if subsequent supplementary corneal topography data shows that the patient has corneal astigmatism but the initial medical record did not record this condition, the system will automatically update the relevant diagnostic content and indicate the reason for the modification. The corrected data is then sent back to the core processing unit.

[0030] The data synchronization module ensures the secure storage and efficient management of medical record data. The cloud storage interface stores medical record data on a cloud server via a distributed database interface, enabling multi-terminal shared access. The encrypted transmission interface uses the AES-256 encryption algorithm to ensure the security of data transmission. The local storage interface caches critical data on local devices to ensure normal system functionality even during network outages. The data backup interface periodically backs up important data to independent storage media to prevent data loss. The core processing unit achieves unified data management and scheduling through the data synchronization module.

[0031] The user interface provides doctors with an intuitive operating environment and a convenient user experience. The visualization unit uses a chart generation interface to transform complex examination data into easily understandable charts, such as visual acuity change curves or corneal topography distribution maps.

[0032] The text rendering interface is responsible for presenting medical record text to users in a clear and aesthetically pleasing format. The operation control unit provides a standardized operation interface, enabling doctors to easily perform operations such as data entry, medical record viewing, and modification. The feedback collection unit records users' operating habits and usage suggestions in real time and transmits this information to the core processing unit for system optimization.

[0033] In practical applications, when a patient comes to the ophthalmology clinic, the nurse first uses a vision testing device to conduct a basic examination. The relevant data is transmitted to the data acquisition module through the vision testing device interface. Then, the doctor observes the patient's eye condition under a slit lamp and inputs the observation results into the system through an unstructured data interface.

[0034] Meanwhile, the corneal topography analysis device transmits the test results to the system via a dedicated interface. All data is processed by the core processing unit, undergoing specialized processing by the semantic parsing module and content optimization by the dynamic correction module, ultimately forming a complete electronic medical record.

[0035] Doctors can view the generated medical records through a user interface and make adjustments to any necessary parts. The system then uses a data synchronization module to store the final version of the medical records in the cloud and synchronize it to the hospital's information management system.

[0036] Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An AI-based eye optometry medical record automatic generation and correction system, characterized in that: The system comprises a core processing unit, the output end of the core processing unit is electrically connected with a data acquisition module, a semantic analysis module, a dynamic correction module, a data synchronization module and a user interaction interface, the output end and the input end of the data acquisition module, the semantic analysis module, the dynamic correction module, the data synchronization module and the user interaction interface are bidirectionally electrically connected with the input end and the output end of the core processing unit.

2. The AI-based eye optometry medical record automatic generation and correction system according to claim 1, characterized in that: The output end of the data acquisition module is electrically connected with a structured data interface, an unstructured data interface and a real-time data stream interface, the input end and the output end of the structured data interface, the unstructured data interface and the real-time data stream interface are bidirectionally electrically connected with the output end and the input end of the data acquisition module, the output end of the structured data interface is electrically connected with a vision detection device interface, a corneal topography analysis interface and an intraocular pressure measurement device interface, the output end and the input end of the vision detection device interface, the corneal topography analysis interface and the intraocular pressure measurement device interface are bidirectionally electrically connected with the input end and the output end of the structured data interface.

3. The AI-based eye optometry medical record automatic generation and correction system according to claim 1, characterized in that: The output end of the semantic analysis module is electrically connected with a medical terminology mapping unit, a context understanding unit and a syntax analysis unit, the input end and the output end of the medical terminology mapping unit, the context understanding unit and the syntax analysis unit are bidirectionally electrically connected with the output end and the input end of the semantic analysis module, the output end of the medical terminology mapping unit is electrically connected with a terminology library interface and a rule engine interface, the input end and the output end of the terminology library interface and the rule engine interface are bidirectionally electrically connected with the output end and the input end of the medical terminology mapping unit.

4. The AI-based eye optometry medical record automatic generation and correction system according to claim 1, characterized in that: The output end of the dynamic correction module is electrically connected with an incremental learning unit, a conflict detection unit and a correction strategy unit, the input end and the output end of the incremental learning unit, the conflict detection unit and the correction strategy unit are bidirectionally electrically connected with the output end and the input end of the dynamic correction module, the output end of the incremental learning unit is electrically connected with a model update interface and a data feedback interface, the input end and the output end of the model update interface and the data feedback interface are bidirectionally electrically connected with the output end and the input end of the incremental learning unit.

5. The AI-based eye optometry medical record automatic generation and correction system according to claim 1, characterized in that: The output end of the data synchronization module is electrically connected with a cloud storage interface, a local storage interface and a data backup interface, the input end and the output end of the cloud storage interface, the local storage interface and the data backup interface are bidirectionally electrically connected with the output end and the input end of the data synchronization module, the output end of the cloud storage interface is electrically connected with a distributed database interface and an encrypted transmission interface, the input end and the output end of the distributed database interface and the encrypted transmission interface are bidirectionally electrically connected with the output end and the input end of the cloud storage interface.

6. The AI-based eye optometry medical record automatic generation and correction system according to claim 1, characterized in that: The output end of the user interaction interface is connected with a visual display unit, an operation control unit and a feedback collection unit, the input end and the output end of the visual display unit, the operation control unit and the feedback collection unit are bidirectionally connected with the output end and the input end of the user interaction interface, the output end of the visual display unit is connected with a chart generation interface and a text rendering interface, and the input end and the output end of the chart generation interface and the text rendering interface are bidirectionally connected with the output end and the input end of the visual display unit.

7. The AI-based eye optometry medical record automatic generation and correction system according to claim 2, characterized in that: The structured data interface is configured to receive standardized data from the visual acuity detection device interface, the corneal topography analysis interface and the tonometry device interface.

8. The AI-based eye optometry medical record automatic generation and correction system according to claim 3, characterized in that: The medical term mapping unit calls a preset ophthalmic and optometric field term library through a term library interface to realize data conversion.

9. The AI-based eye optometry medical record automatic generation and correction system according to claim 4, characterized in that: The incremental learning unit receives the latest training data through a model updating interface and updates the internal algorithm online in combination with the information provided by the data feedback interface.

10. The AI-based eye optometry medical record automatic generation and correction system according to claim 5, characterized in that: The cloud storage interface stores data to a cloud server through a distributed database interface and realizes safe transmission of the data through an encrypted transmission interface.

Images