Medical image integration platform and method based on AI

Abstract

The invention relates to an AI-based medical image integration platform and method, and the platform is integrated in a PACS system, and comprises an AI engine integration module, a data management module and a user interface module which are connected in sequence. The AI engine integration module is used for sending medical images of the PACS system to a plurality of AI systems for respective processing, quickly integrating data processed by the AI systems through a standard API interface, sending the integrated data to the data management module for storage, and then transmitting the integrated data to the user interface module for display and interaction of image data. Compared with the prior art, the system has the advantages of integrating different AI systems, enhancing the security of medical image data transmission, enhancing the real-time interaction function and the like.

Description

Technical Field

[0001] The present invention relates to the field of medical information systems, and in particular to an AI-based medical imaging integration platform and method. Background Art

[0002] In modern medicine, the Picture Archiving and Communication System (PACS) is an important part of the hospital's imaging department. The PACS system is responsible for the acquisition, transmission, storage and management of image data, which can improve the hospital's image processing efficiency. However, with the development of medical AI technology, more and more medical institutions have introduced AI systems for image analysis and diagnosis. These AI systems can extract valuable information from image data through technologies such as deep learning to assist doctors in diagnosis.

[0003] At present, there are some problems in the integration between the existing PACS system and the AI ​​system. The main problems include:

[0004] a) The data formats and interface standards of different AI systems are inconsistent, resulting in high integration complexity;

[0005] b) There are security and privacy risks during data transmission and processing;

[0006] c) Lack of real-time interactive functions, doctors cannot easily obtain and feedback AI analysis results.

[0007] Based on the above main problems, the processing efficiency of medical imaging data is low, there are certain security risks, and doctor interaction is inconvenient.

[0008] How to achieve the integration of medical imaging data with different AI interface standards while improving security and interactivity has become a technical problem that needs to be solved. Summary of the invention

[0009] The purpose of the present invention is to provide an AI-based medical imaging integration platform and method in order to overcome the defects of the above-mentioned prior art.

[0010] The purpose of the present invention can be achieved by the following technical solutions:

[0011] According to one aspect of the present invention, there is provided an AI-based medical imaging integration platform, which is integrated into a PACS system and includes an AI engine integration module, a data management module, and a user interface module connected in sequence;

[0012] The AI ​​engine integration module sends the medical images of the PACS system to multiple AI systems for separate processing, and quickly integrates the data processed by the AI ​​systems through a standard API interface, sends the integrated data to the data management module for encrypted, compressed storage, and then transmits it to the user interface module for display and interaction of the image data.

[0013] Preferably, the standard API interface of the AI ​​engine integration module provides a unified data format and communication protocol, supports the rapid integration of multiple AI systems, wherein the rapid integration is achieved through a plug-in mechanism and allows dynamic loading and updating of different AI algorithms.

[0014] More preferably, the AI ​​engine integration module supports conversion of data in different formats into a standard format, and conversion between different data formats.

[0015] Preferably, the AI ​​engine integration module supports parallel processing of AI analysis results, and performs fusion and comparison of AI analysis results according to the needs of doctors.

[0016] Preferably, the data management module adopts Hadoop distributed file system as a data storage solution, supports DICOM protocol to transmit medical image data between devices, and uses image compression algorithm to perform compression and decompression operations before and after the transmission of medical image data.

[0017] Preferably, the data management module uses the AES algorithm to encrypt and store medical imaging data and manage permissions, and uses the SSL or TLS protocol to ensure the security of data transmission.

[0018] Preferably, the user interface module includes a report generation and feedback unit, a 2D, 3D reconstruction and multi-planar reformation unit, and an AI analysis and result integration unit;

[0019] The 2D, 3D reconstruction and multi-planar reorganization unit is used for 2D, 3D rendering and multi-view synchronous display of various image data, and supports simultaneous viewing of multiple image slices and reconstructed images, and comparison and analysis;

[0020] The AI ​​analysis and result integration unit is used to integrate the API interfaces of different AI systems, comprehensively process the analysis results of multiple AI systems, and uniformly display the analysis results on one interface;

[0021] The report generation and feedback unit is used to integrate AI analysis results with electronic medical records and laboratory results, and adopts a modular design to automatically generate a comprehensive diagnostic report.

[0022] More preferably, the interaction includes:

[0023] The AI ​​analysis and result integration unit supports a real-time feedback function, and the doctor annotates and modifies the AI ​​analysis results and transmits the feedback information back to the AI ​​system;

[0024] The report generation and feedback unit feeds back the doctor's review or modification opinions on the automatically generated comprehensive diagnosis report to the AI ​​system.

[0025] Preferably, the platform also includes a preprocessing module connected to the AI ​​engine integration module, and the preprocessing module automatically selects the most suitable image processing algorithm according to different types of image data through machine learning technology, including denoising, enhancement and standardization.

[0026] According to another aspect of the present invention, a medical image integration method based on AI is provided, the method comprising the following steps:

[0027] Step S1, preprocessing the acquired medical image data, wherein the preprocessing includes denoising, enhancement and standardization operations;

[0028] Step S2, sending the preprocessed results to the corresponding AI system for analysis according to the settings;

[0029] Step S3, the AI ​​system feeds back the analysis results, integrates the analysis results from different AI systems using a standard API interface, and encrypts and compresses them for storage;

[0030] In step S4, the integrated results are integrated, rendered and displayed in 2D and 3D on the user interface, and a comprehensive diagnosis report is automatically generated. The doctor interacts with the AI ​​system through the user interface.

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] 1) The present invention quickly integrates the analysis results of multiple AI systems through a standardized API interface, solving the problem of different formats and communication protocols of different AI systems; then, a variety of image data are displayed and interacted on the user interface, improving the real-time interaction function, and improving the AI ​​system model based on feedback information, thereby helping doctors improve diagnostic efficiency and accuracy.

[0033] 2) The present invention encrypts, compresses, stores and manages permissions of the integrated data, thereby improving the security and privacy of the data during transmission and processing.

[0034] 3) The present invention realizes 2D, 3D rendering and multi-view synchronous display through a customizable user interface, supports simultaneous viewing of multiple image slices and reconstructed images, and performs comparison and analysis; integrates AI analysis results, electronic medical records and laboratory results, adopts a modular design, automatically generates comprehensive diagnostic reports, and improves the usability of AI analysis results. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 This is a schematic diagram of the structure of the AI-based medical imaging integration platform of the present invention;

[0036] Figure 2 It is a flowchart of the medical image integration method based on AI of the present invention;

[0037] In the accompanying drawings, 11 is a preprocessing module, 12 is an AI engine integration module, 13 is a data management module, 14 is a user interface module, 141 is a report generation and feedback unit, 142 is a 2D, 3D reconstruction and multi-planar restructuring unit, 143 is an AI analysis and result integration unit, and 15 is a system management module. DETAILED DESCRIPTION

[0038] 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 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 should fall within the scope of protection of the present invention.

[0039] In order to solve the above problems, the present invention provides an AI-based medical imaging integration platform, which can achieve deep integration of the PACS system and the AI ​​system, improve the processing efficiency and security of image data through standardized interfaces and business processes, and provide real-time interactive functions.

[0040] Example 1

[0041] This embodiment relates to an AI-based medical imaging integration platform, which includes:

[0042] 1) Preprocessing module 11

[0043] In order to improve the analysis accuracy of the AI ​​system, the present invention integrates an automated image preprocessing module into the PACS system. This module uses advanced image processing algorithms, including denoising, enhancement, and standardization steps, to ensure the consistency and high quality of image data. Through machine learning technology, the preprocessing module can automatically select the most suitable preprocessing algorithm according to different types of image data (such as X-ray, CT, MRI, etc.), thereby improving the preprocessing effect.

[0044] In addition, the pre-processing module also supports multi-level processing of image data, including primary processing (such as denoising) and advanced processing (such as texture enhancement), further improving the clarity and detail of the image. This multi-level processing method not only improves the analysis accuracy of the AI ​​system, but also significantly reduces the workload of doctors.

[0045] 2) AI engine integration module 12: This module provides a standardized API interface, and supports rapid integration with a variety of AI products through a unified data format and communication protocol. The core of this module is its flexible plug-in mechanism, which allows dynamic loading and updating of different AI algorithms, ensuring that the platform can quickly adapt to the latest developments in AI technology. The integration module not only supports parallel processing of AI analysis results, but also can fuse and compare results according to the needs of doctors, improving the accuracy and efficiency of doctors' diagnoses.

[0046] This module includes plugin registration management, where developers can register new 3D reconstruction plugins through defined interfaces. Each plugin provides metadata, version information, and dependencies when registering, and the platform manages and schedules the plugins based on this information.

[0047] At runtime, the integration platform dynamically loads the required reconstruction plug-ins through the plug-in management system. Using technologies such as Java's SPI (Service Provider Interface) mechanism or Python's entry point mechanism (entry points), the platform can select and load appropriate plug-ins as needed.

[0048] The client reads the configuration file and embeds or opens the front-end tool provided by the AI ​​vendor or opens a new interface as needed to load and use the corresponding plug-in. Use the configuration file to define the plug-in information (such as path, version, and dependencies), and the client loads the plug-in according to the configuration file. The metadata of the plug-in is defined in the configuration file, including the plug-in name, version, path, and dependencies.

[0049] Metadata definition: Create a standard metadata format that uses JSON or XML to describe plugin information.

[0050] Metadata management tool: Develop a tool or service that is responsible for reading, parsing, and managing the plugin's metadata. You can use a registry or configuration management tool to store and retrieve this metadata.

[0051] Each AI function module and plugin should contain a description file (such as 'metadata.json' or 'metadata.xml') to store metadata, version information, and dependencies. Metadata can include plugin name, version number, vendor information, dependencies (other required plugins or libraries), supported functions, etc.

[0052] The integration platform supports version control and update mechanisms for plugins. Through version control systems (such as Git) and continuous integration / continuous deployment (CI / CD) pipelines, the platform can automatically deploy the latest versions of plugins to ensure that the reconstruction technology is always at the forefront.

[0053] The inconsistent data formats and interface standards of different AI systems lead to high integration complexity. The implementation methods include:

[0054] i) Standardized interface: Each AI function module provides a unified interface and supports two data formats, XML and JSON. This ensures compatibility and flexibility between different systems.

[0055] ii) Data format conversion: A data format conversion module is implemented in the integration platform, which is responsible for converting data in different formats into a standard format.

[0056] Specific implementation steps:

[0057] Interface standardization: Develop unified API standards, and all AI modules must implement these standard interfaces. The interface definition should include support for data formats (JSON and XML) and corresponding request / response specifications.

[0058] Data format conversion module: Implement a data conversion module responsible for converting between different data formats. Ensure that the platform can receive and process data in different formats.

[0059] API interface definition: Each AI module provides a unified analysis interface, such as ' / api / analyze', which supports POST requests and accepts data in JSON or XML format. The interface returns the analysis results in JSON or XML format, depending on the client's request header (Content-Type).

[0060] Assuming there are AI systems A and B, the integration platform calls different AI modules through standardized interfaces and automatically handles data format conversion based on the request header.

[0061] The integration platform provides RESTful API and GraphQL interfaces, using JSON or XML format for data exchange. The API interface design follows standardized specifications to ensure that different AI products can be quickly integrated. The platform manages and routes API requests through the API gateway to ensure efficient data exchange and processing.

[0062] The implementation methods for data exchange using JSON and XML formats include:

[0063] Designed RESTful API and SOAP WebService, using JSON and XML as data exchange formats respectively.

[0064] RESTful API uses JSON format for data exchange and is suitable for lightweight, frequently called services.

[0065] SOAP WebService uses XML format for data exchange and is suitable for complex data transmission that requires strict protocols.

[0066] The specific implementation steps for data exchange using JSON and XML formats include:

[0067] Design API interface: define API endpoints, request methods (GET, POST, PUT, DELETE), and JSON formats for requests and responses.

[0068] Implement WebService: Define the WSDL file of SOAP WebService, describe the service interface, message format and communication protocol, and use XML Schema to define the data structure.

[0069] To ensure that different AI products can be quickly integrated, use an enterprise service bus (ESB) or middleware to provide standardized interfaces and protocols so that different AI products can be quickly integrated into the platform. Use unified communication protocols (such as HTTP, gRPC) and data formats (such as JSON, XML) to ensure data transmission and interoperability between different AI modules. Deploy ESB, such as Apache Camel or Mule ESB, as an intermediate layer to handle data exchange and service calls of different AI products. Define standard interfaces: Define and manage service interfaces through ESB, standardize the access standards of AI products, and simplify the integration process.

[0070] 3) Data management module 13: This module uses distributed storage technology and efficient image compression algorithms to ensure the rapid storage and transmission of large-scale medical imaging data. By supporting the DICOM standard, this module can be seamlessly compatible with various medical imaging devices to ensure the consistency and high quality of imaging data.

[0071] In addition, the data management module 13 also integrates advanced data encryption and authority management technologies to ensure the security and privacy of image data during transmission and storage.

[0072] Distributed storage technology: This platform uses Hadoop Distributed File System (HDFS) as a data storage solution to distribute and store large-scale medical imaging data on multiple server nodes. Each node stores a portion of the data and ensures data reliability and availability through redundancy and copy mechanisms. During data storage, HDFS automatically manages the location and replication of data blocks to ensure that the platform can continue to operate in the event of any single point of failure. The distributed computing framework enables the system to process data in parallel to ensure real-time reconstruction and analysis. The distributed computing framework uses a task scheduling system (such as Apache Airflow, Spring Batch) to manage and schedule distributed computing tasks. Large data sets are divided into multiple small data blocks and distributed on different computing nodes for processing. Use the distributed computing framework to perform data processing tasks and use cluster resources to improve computing speed and efficiency. Define and manage the dependencies and scheduling strategies of computing tasks to ensure the orderly execution of tasks and the merging of results.

[0073] In addition, the platform also integrates GPU acceleration technology, using technologies such as CUDA and OpenCL to improve data processing performance.

[0074] Efficient image compression algorithm: The platform uses efficient image compression algorithms such as JPEG 2000 and HEVC (High Efficiency Video Coding). JPEG 2000 is an image compression standard based on wavelet transform, which can achieve high compression ratio without obvious quality loss. HEVC is used to compress video data and improve transmission efficiency. Image data will be compressed before transmission and decompressed at the receiving end to ensure efficient transmission and data integrity.

[0075] Support for DICOM standards: The platform is fully compatible with the DICOM (Digital Imaging and Communications in Medicine) standard and supports standardized storage, transmission and display of image data. Through the DICOM protocol, image data can be transmitted between devices with consistency and integrity. The platform provides DICOM server and client functions to ensure seamless integration with a variety of medical imaging devices.

[0076] Data encryption and rights management: Data encryption uses the AES (Advanced Encryption Standard) algorithm to encrypt and store image data, and uses the SSL / TLS protocol to ensure the security of data transmission. Rights management uses technologies such as OAuth 2.0 and LDAP (Lightweight Directory Access Protocol) to achieve fine-grained access control. When users access image data, they need to perform identity authentication and rights checks to ensure that only authorized users can access and operate sensitive data, thereby ensuring data security and privacy.

[0077] 4) User interface module 14: The user interface module 14 is designed with a friendly and highly customizable interface, and doctors can adjust the operation interface according to their personal habits. Users can personalize the interface through configuration files or settings to meet the usage habits of different users.

[0078] The user interface module 14 includes a report generation and feedback unit 141: In order to simplify the doctor's workflow, the present invention develops an integrated report generation function in the PACS system. This function supports the integration of image analysis results with other medical data (such as electronic medical records, laboratory results, etc.) to automatically generate a comprehensive diagnostic report. The report generation module adopts a templated design to automatically generate corresponding report templates according to different types of diseases and diagnostic needs, thereby improving the efficiency of report generation. The generated report content is detailed and logical, including image analysis results, clinical diagnostic opinions and suggestions, etc. Doctors can review and modify the automatically generated reports, and feedback the modification opinions to the AI ​​system to continuously improve the analysis ability and accuracy of the AI ​​system. In addition, the report generation module also supports multiple output formats (such as PDF, DICOM SR), which is convenient for doctors to archive and share.

[0079] The user interface module 14 supports multiple display modes of image data, including 2D, 3D reconstruction and a multi-planar reformat unit (MPR) 142 .

[0080] The user interface module 14 adopts a modular design, and each functional module (such as 2D viewer, 3D reconstruction viewer, MPR viewer) can be loaded and run independently.

[0081] In addition, the user interface module 14 also integrates advanced interactive technologies, such as voice control and gesture recognition, making the doctor's operation more convenient and efficient.

[0082] Friendly and highly customizable interface: Use modern front-end frameworks such as React, Angular, or Vue to design modular and customizable user interfaces. Users can adjust the operation interface through configuration files or settings to meet personal habits and workflow requirements. The interface design follows human-computer interaction (HCI) best practices to ensure ease of use and intuitiveness of the interface.

[0083] Multiple ways to display image data: 2D, 3D reconstruction and multi-planar reformation (MPR) functions are realized, and 3D rendering and display are performed using technologies such as WebGL or Three.js. The user interface supports multi-view synchronous display, and users can view multiple image slices and reconstructed images at the same time for comparison and analysis. The platform provides image zoom, rotation, translation and other operation tools, so users can easily operate and view image data.

[0084] Advanced interactive technology: Integrates voice control and gesture recognition technology, uses Web Speech API and LeapMotion and other technologies to achieve natural human-computer interaction. Users can control interface operations through voice commands, or operate and analyze image data through gesture recognition, improving the convenience and efficiency of operations.

[0085] The user interface module 14 calls the reconstruction service of the backend through the API and displays the processing results to the user in real time.

[0086] The user interface module 14 also includes an AI analysis and result integration unit 143: This module integrates the API interfaces of different AI manufacturers and displays various AI analysis results in a unified interface. Through a standardized data interface, the analysis results of different AI systems can be seamlessly integrated, reducing the complexity of data conversion and processing. In order to facilitate doctors to compare and screen, the AI ​​analysis module provides a variety of views and screening functions, and doctors can choose to view the analysis results of different AI systems according to different needs.

[0087] In addition, the AI ​​analysis module also supports real-time feedback function. Doctors can annotate and modify the AI ​​analysis results. The system will send this feedback information back to the AI ​​model to continuously optimize and improve the analysis capabilities of the AI ​​system.

[0088] The module also has a result integration function, which uses advanced fusion algorithms to comprehensively process the analysis results of multiple AI systems. The fusion algorithm uses a machine learning model to perform weighted averaging or rule-based comprehensive processing on the analysis results of multiple AI systems to provide more comprehensive and accurate diagnostic recommendations.

[0089] 5) System Management Module 15: System Management Module 15 integrates comprehensive system monitoring and maintenance functions. Through real-time performance monitoring and fault diagnosis, the platform can quickly identify and handle potential problems to ensure the stable operation of the platform. The module also provides detailed logging and auditing functions, supports the tracing and analysis of user operations and system events, and further improves the security of the system.

[0090] Comprehensive system monitoring and maintenance functions: Integrate open source monitoring tools such as Prometheus and Grafana for real-time performance monitoring and fault diagnosis. System monitoring includes monitoring of resource usage such as CPU, memory, disk, network, and the running status and performance indicators of applications. The fault diagnosis module can automatically detect and identify system faults, trigger alarm mechanisms, and provide detailed fault reports and solutions.

[0091] Detailed logging and auditing functions: Use ELK (Elasticsearch, Logstash, Kibana) and other log management systems to record all operation logs and system events. Log records include detailed information such as user login, data access, system operation, and provide real-time log query and analysis functions. The audit function analyzes log data and generates audit reports to support the tracing and analysis of user operations and system events, thereby improving system security.

[0092] Example 2

[0093] This embodiment also relates to an AI-based medical image integration method, which includes:

[0094] Step S1, preprocessing the acquired medical image data to ensure the consistency and high quality of the image data, including denoising, enhancement and standardization operations;

[0095] Step S2, sending the preprocessed results to the corresponding AI system for analysis according to the settings;

[0096] Step S3, the AI ​​system feeds back the analysis results, and the analysis results from different AI systems are integrated using a standard interface and stored in the data management module 13;

[0097] In step S4, the integrated results are integrated, rendered and displayed in 2D and 3D on the user interface, and a comprehensive diagnostic report is automatically generated. The doctor interacts with the AI ​​system through the user interface, including: the doctor annotates and modifies the AI ​​analysis results and sends this feedback information back to the AI ​​system; the doctor's review or modification opinions on the automatically generated comprehensive diagnostic report are fed back to the AI ​​system.

[0098] After receiving feedback from doctors, the AI ​​system improves its model.

[0099] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. An AI-based medical imaging integration platform, characterized in that: The platform is integrated into the PACS system, and includes an AI engine integration module (12), a data management module (13), and a user interface module (14) which are connected in sequence; The AI ​​engine integration module (12) sends the medical images of the PACS system to multiple AI systems for separate processing, and quickly integrates the data processed by the AI ​​systems through a standard API interface, and sends the integrated data to the data management module (13) for encrypted, compressed storage, and then transmits it to the user interface module (14) for display and interaction of the image data.

2. The AI-based medical imaging integration platform according to claim 1, characterized in that: The standard API interface of the AI ​​engine integration module (12) provides a unified data format and communication protocol, supports the rapid integration of multiple AI systems, wherein the rapid integration is achieved through a plug-in mechanism and allows dynamic loading and updating of different AI algorithms.

3. The AI-based medical imaging integration platform according to claim 2, characterized in that: The AI ​​engine integration module (12) supports conversion of data in different formats into a standard format, and conversion between different data formats.

4. The AI-based medical imaging integration platform according to claim 1, characterized in that: The AI ​​engine integration module (12) supports parallel processing of AI analysis results and integrates and compares AI analysis results according to the needs of doctors.

5. The AI-based medical imaging integration platform according to claim 1, characterized in that: The data management module (13) adopts the Hadoop distributed file system as a data storage solution, supports the DICOM protocol to transmit medical image data between devices, and adopts an image compression algorithm to perform compression and decompression operations before and after the transmission of the medical image data.

6. The AI-based medical imaging integration platform according to claim 1, characterized in that: The data management module (13) adopts the AES algorithm to encrypt and store the medical image data and manage the rights, and uses the SSL or TLS protocol to ensure the security of data transmission.

7. The AI-based medical imaging integration platform according to claim 1, characterized in that: The user interface module (14) includes a report generation and feedback unit (141), a 2D, 3D reconstruction and multi-planar reconfiguration unit (142), and an AI analysis and result integration unit (143); The 2D, 3D reconstruction and multi-planar reorganization unit (142) is used for two-dimensional, three-dimensional rendering and multi-view synchronous display of multiple image data, and supports simultaneous viewing of multiple image slices and reconstructed images, and comparison and analysis; The AI ​​analysis and result integration unit (143) is used to integrate the API interfaces of different AI systems, comprehensively process the analysis results of multiple AI systems, and uniformly display the analysis results on one interface; The report generation and feedback unit (141) is used to integrate AI analysis results with electronic medical records and laboratory results, and adopts a modular design to automatically generate a comprehensive diagnosis report.

8. The AI-based medical imaging integration platform according to claim 7, characterized in that: The interaction includes: The AI ​​analysis and result integration unit (143) supports a real-time feedback function, and the doctor annotates and modifies the AI ​​analysis results and transmits the feedback information back to the AI ​​system; The report generation and feedback unit (141) feeds back the doctor's review or modification opinions on the automatically generated comprehensive diagnosis report to the AI ​​system.

9. The AI-based medical imaging integration platform according to claim 1, characterized in that: The platform also includes a preprocessing module (11) connected to the AI ​​engine integration module (12), and the preprocessing module (11) automatically selects the most suitable image processing algorithm according to different types of image data through machine learning technology, including denoising, enhancement and standardization.

10. A method for using the AI-based medical imaging integration platform according to any one of claims 1 to 9, characterized in that: The method comprises the following steps: Step S1, preprocessing the acquired medical image data, wherein the preprocessing includes denoising, enhancement and standardization operations; Step S2, sending the preprocessed results to the corresponding AI system for analysis according to the settings; Step S3, the AI ​​system feeds back the analysis results, integrates the analysis results from different AI systems using a standard API interface, and encrypts and compresses them for storage; In step S4, the integrated results are integrated, rendered and displayed in 2D and 3D on the user interface, and a comprehensive diagnosis report is automatically generated. The doctor interacts with the AI ​​system through the user interface.

Images