Medical Document Creation Support Apparatus for Readable Reports
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies for creating medical documents from images with multiple regions of interest generate lengthy lists of sentences, making the documents difficult to read and not effectively supporting document creation.
Innovation Solution
A document creation support apparatus that acquires information on multiple regions of interest in a medical image, derives an evaluation index for each, and generates text based on these indices to prioritize and format the information, determining which regions to include, the level of detail, and the format (sentence, bullet, or tabular) for improved readability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If sentences are generated for each individual region of interest and listed in the medical document, then all regions of interest are documented, but the medical document becomes difficult to read
Solution Approach 1:
The patent combines multiple individual region descriptions into a unified, integrated text format. Instead of listing separate sentences for each region of interest, the system merges them into a coherent narrative that maintains all essential information while improving readability through proper sentence structure and logical flow.
Solution Approach 2:
The patent applies different description styles and levels of detail to different regions of interest based on their characteristics and importance. Each region is described with appropriate local quality - some regions receive more detailed descriptions while others are summarized, creating a balanced document that is both comprehensive and readable.
2Loss of information
If multiple regions of interest are described in detail, then comprehensive documentation is achieved, but the document length increases and becomes less readable
Solution Approach 1:
The patent implements partial description strategy where not all regions of interest are described with the same level of detail. Based on evaluation indices, the system selectively describes only the most important regions in detail while providing brief mentions or summaries for less critical regions, thereby reducing overall document length while maintaining essential information.
Solution Approach 2:
The patent extracts and prioritizes only the most essential information from multiple regions of interest. By using evaluation indices to identify critical regions and features, the system extracts only the necessary details for each region, eliminating redundant information and creating a more concise document that maintains completeness of essential findings.
3Loss of information
If all features of all regions of interest are included in the text, then complete documentation is achieved, but the text becomes overly detailed and difficult to read
Solution Approach 1:
The patent applies local quality by describing different features of different regions with appropriate levels of detail. Each region's description is tailored to its specific characteristics and clinical importance, including only the features that are relevant and necessary for that particular region, thereby reducing overall text complexity while maintaining feature completeness.
Solution Approach 2:
The patent implements partial description of features by selectively including only the most important characteristics of each region of interest. Based on evaluation indices, the system determines which features warrant detailed description and which can be summarized or omitted, creating a balanced text that is both comprehensive and readable.
Data Source
AI summary
A document creation support apparatus acquires information indicating a plurality of regions of interest included in a medical image, derives an evaluation index as a target of a medical document for each of the plurality of regions of interest, and generates text including a description regarding at least one of the plurality of regions of interest based on the evaluation index.