Link Character Insertion for Medical Image Lesion Positioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image interpretation report generation methods require complex linking operations and fail to accurately represent the position of lesion regions in medical images, increasing workload and making it difficult to understand the corresponding medical image.
Innovation Solution
An image interpretation report generation apparatus and method that allows for easy insertion of link characters by inputting the position of lesion regions, with the option to automatically detect these regions, and displays an index representing the position, enabling distinguishable display of link characters and corresponding medical image thumbnails.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If a linking operation is performed by dragging and dropping reference images to character strings, then hyperlinks can be generated in image interpretation reports, but the work load in input increases
Solution Approach 1:
The system automatically generates hyperlinks by detecting lesion regions in medical images and creating link characters without requiring manual dragging and dropping operations. The apparatus self-services the hyperlink generation process, eliminating the need for user intervention in the linking operation.
Solution Approach 2:
The system performs preliminary detection of lesion regions and pre-generation of link characters before the final report assembly. By preparing the linking infrastructure in advance through automatic lesion region detection, the system eliminates the need for manual linking operations during report generation.
2Extent of automation
If a character string is linked to an image, then hyperlinks can be created, but when plural link target regions are present in a slice image, only the same corresponding medical image is observable, making it impossible to recognize the position indicated by each character string
Solution Approach 1:
The system segments the medical image into multiple region-of-interest (ROI) images based on detected lesion regions. Each link character is associated with a specific ROI image, allowing users to distinguish and access individual lesion positions. This segmentation preserves position information by creating separate observable images for each link target region.
Solution Approach 2:
The system assigns local quality attributes to different link characters by associating them with specific ROI images. Each link character has unique properties (such as color coding or positioning) that correspond to its associated lesion region, enabling users to recognize and distinguish the position indicated by each character string.
3Measurement precision
If new plural reference images are generated by cutting out lesion regions, then position recognition is improved, but the edit operation becomes extremely complicated
Solution Approach 1:
The apparatus automatically performs the cutting out of lesion regions and generation of ROI images without requiring manual edit operations. The system self-services the image processing tasks, including automatic lesion region detection, ROI extraction, and link character generation, thereby avoiding complicated manual editing while maintaining position recognition accuracy.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
[Aim] To assist an operation for linking an image interpretation report and a medical image to each other [Composition] A medical image is obtained and displayed. An electronic image interpretation report of the medical image is generated and displayed. The position of a lesion region in the medical image is input. A link character linked by a hyperlink to the medical image including the input position of the lesion region is inserted into the image interpretation report in such a manner that the link character is distinguishably displayed. The distinguishably displayed link character in the image interpretation report is selected. An index representing the position of the lesion region is displayed in the medical image based on the selection of the link character.