Medical Image Report Grouping to Reduce Radiologist Review Burden
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The burden on radiologists increases due to excessive information and character count in interpretation reports generated for medical images with multiple abnormal shadows, making it difficult to efficiently review the contents.
Innovation Solution
A document creation support apparatus that generates first texts for individual feature portions, classifies them into groups, and integrates these texts into second and third texts, allowing for summarized reporting and easy access to detailed information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If a report is generated for each abnormal shadow in a medical image, then the completeness of information is improved, but the character count and information amount becomes excessive
Solution Approach 1:
The patent segments the report generation process into two levels: individual abnormal shadow reports and grouped abnormal shadow reports. The processor generates reports for each abnormal shadow separately, then groups them by anatomical location or type, creating a hierarchical structure that preserves complete information while organizing it in a manageable format.
Solution Approach 2:
The patent merges multiple individual abnormal shadow reports into a consolidated report by grouping abnormal shadows with similar characteristics or locations. This combining process reduces the overall character count by eliminating redundant information while maintaining the essential details of each abnormal shadow through structured presentation.
2Measurement precision
If a report is generated for each abnormal shadow in a medical image, then the detail level is improved, but the review burden increases
Solution Approach 1:
The patent segments the review process by allowing radiologists to first review a summary view of grouped abnormal shadows, then drill down into individual abnormal shadow details only when needed. This segmented approach reduces the initial review burden while preserving access to complete detailed information.
Solution Approach 2:
The patent implements a dynamic report structure where the presentation format can be adjusted based on the number of abnormal shadows and the radiologist's review needs. The system can dynamically switch between condensed grouped views and expanded individual views, making the review process more adaptable and less burdensome.
3Loss of information
If the number of abnormal shadows increases in a medical image, then the comprehensiveness of the report is improved, but the character count increases excessively
Solution Approach 1:
The patent changes the organizational parameters of the report based on the number of abnormal shadows detected. When the number of abnormal shadows is large, the system automatically applies grouping parameters and summary statistics, transforming the report structure from a simple list of individual findings to a hierarchical organization that reduces character count while maintaining comprehensiveness.
Solution Approach 2:
The patent uses template-based report generation where common patterns and repeated information are captured in reusable templates. This allows the system to generate comprehensive reports for multiple abnormal shadows without proportionally increasing the character count, as similar findings can be referenced through template instances rather than fully repeated.
Data Source
AI summary
A document creation support apparatus comprises at least one processor. The processor is configured to generate a first text describing a property of a feature portion for each of a plurality of feature portions included in an image, classify the plurality of feature portions into at least one group, generate a second text integrating the first text generated for each of the feature portions belonging to the same group for each group, and generate a third text integrating the second text generated for each group.