Medical Image Generation With Analysis-Basis Visualization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The reliability of analysis results obtained using machine models for medical image analysis is often unclear due to a lack of transparency in the basis for the derived results.
Innovation Solution
An image generation device and method that includes acquiring an analysis result and derivation basis data from a machine model, and generating a display image by changing the medical image based on this data, allowing both the result and its basis to be displayed together.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If machine model analysis is used to analyze medical images, then analysis efficiency and accuracy are improved, but the transparency and reliability of the analysis basis become unclear
Solution Approach 1:
The patent extracts the derivation basis data from the machine model's internal processing and presents it separately as a key component. The derivation basis data indicates the specific regions or features in the medical image that the machine model used for analysis, making the previously opaque analysis basis transparent and verifiable.
Solution Approach 2:
The patent introduces derivation basis data as an intermediary element between the machine model and the final analysis result. This intermediary component bridges the gap by showing users not only the analysis result but also the specific image regions that formed the basis for that result, enabling traceability and understanding.
2Ease of operation
If only analysis results are displayed, then display simplicity is maintained, but user trust and understanding of the analysis basis are reduced
Solution Approach 1:
The patent merges the display of analysis results with the display of derivation basis data in a unified interface. The medical image is displayed with overlays or annotations that simultaneously show both the analysis results and the corresponding derivation basis regions, allowing users to view both components together without increasing complexity.
Solution Approach 2:
The patent uses color coding or visual differentiation to highlight derivation basis regions in the medical image. By applying distinct colors or visual markers to regions that serve as analysis basis, the system makes the derivation basis visually distinguishable while maintaining overall display simplicity and user-friendly navigation.
3Reliability
If derivation basis data is added to the display, then analysis reliability is improved, but display complexity increases
Solution Approach 1:
The patent segments the medical image into distinct regions: the original image areas and the derivation basis regions. By segmenting the display into these clear, distinguishable parts with different visual characteristics, the system presents complex derivation basis data in an organized manner that does not overwhelm the user.
Solution Approach 2:
The patent adds derivation basis information as an additional dimensional layer over the medical image, rather than presenting it as a separate complex interface. This dimensional approach allows the derivation basis data to be overlaid on the image in a way that provides context without requiring users to navigate through separate complex data structures.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
The reliability of an analysis result obtained using a machine model is further improved. An image generation device includes an acquirer that acquires an analysis result output from a machine model that analyzes a medical image obtained by imaging a subject and derivation basis data indicating a basis on which the analysis result is derived, and an image generator that generates a display image obtained by changing the medical image based on the derivation basis data.