Endoscopy Support for AI Region Scoring and Exam Data Linking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional endoscopy systems fail to organize and manage large amounts of image information and analysis results efficiently, especially when used by multiple users, making it difficult for users to utilize the data effectively.
Innovation Solution
An endoscopy support apparatus and method that includes units for generating analysis, user, and examination management information by associating image-related information with examination periods, using AI models to estimate target regions and score likelihood, and user interface inputs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If image information and analysis results are stored in a storage device in a conventional endoscopy system, then the data can be retained for future use, but the large amount of image information and analysis results are not organized and are stored in a complicated manner
Solution Approach 1:
The patent segments the storage system by creating separate functional units: an analysis information generation unit that processes images during examination, an image-related information generation unit that organizes images with metadata, and an examination management information generation unit that structures data by examination ID. This segmentation transforms the complicated unorganized storage into a structured hierarchical system where data is divided into manageable organized components.
Solution Approach 2:
The patent applies preliminary action by organizing and structuring image data and analysis results during the examination process itself, rather than storing them raw and organizing later. The system pre-associates image information with examination IDs, user information, and analysis results through the image-related information generation unit and examination management information generation unit, making the data ready for efficient retrieval and use without requiring post-examination organization.
2Productivity
If the endoscopy system is used continuously by a plurality of users, then the system can serve multiple users efficiently, but it becomes difficult to organize and manage the large amount of image information obtained during examination periods for each user
Solution Approach 1:
The patent introduces intermediary structures to manage multi-user data: examination management information that acts as a mediator between multiple users and the image database, and image-related information that serves as an intermediary layer between raw images and user access. These intermediary structures organize images by examination ID and associate them with user information, enabling efficient multi-user access while maintaining clear data management boundaries.
Solution Approach 2:
The patent adds organizational dimensions to the data structure by introducing examination ID as a new dimension for organizing images, in addition to user information. This dimensional change allows the system to manage data along multiple axes (user, examination, time) simultaneously, enabling efficient retrieval and management of large volumes of multi-user examination data without increasing operational complexity.
3Loss of information
If image analysis is performed on each piece of image information during examination, then useful information can be extracted in real time, but the results are not properly organized and associated with the corresponding images and examination contexts
Solution Approach 1:
The patent merges previously separate functions into an integrated system: the analysis information generation unit combines image processing with analysis result generation, the image-related information generation unit combines images with metadata and analysis results, and the examination management information generation unit combines all elements under examination ID. This merging eliminates the information loss problem by ensuring all data elements are properly associated while managing complexity through functional integration rather than separate discrete components.
Data Source
AI summary
An endoscopy support apparatus 1 includes: an analysis information generation unit 2 that inputs image information of an imaged living body into a model, estimates a region of a target site of the living body, and generates analysis information including region information indicating the estimated region and score information indicating likeness of the region to the target site; a user information generation unit 3 that generates user information related to the user, which has been input by a user using a user interface 23; an image-related information generation unit 4 that generates image-related information by associating imaging date and time information, the analysis information, and the user information, for each piece of image information; and an examination management information generation unit 5 that generates examination management information by associating a plurality of pieces of the image-related information with examination information indicating an examination period.