Medical Image Lesion Region Selection Using Evaluation Values
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing medical image processing techniques present all regions including lesions to users, which can reduce efficiency by including regions of relatively low importance.
Innovation Solution
An information processing apparatus that detects regions of body parts and lesions from medical images, derives an evaluation value for each region based on the degree of certainty of the lesion, and selects regions with evaluation values equal to or greater than a threshold value, allowing for appropriate selection of regions including lesions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If all regions including lesions are presented to the user, then complete information is provided, but interpretation efficiency is reduced due to inclusion of low-importance regions
Solution Approach 1:
The patent changes the parameter of region selection by introducing an evaluation value that combines multiple factors (lesion size, degree of certainty, anatomical importance) to filter and prioritize regions. This allows the system to maintain completeness of important lesion information while excluding low-importance regions, thereby improving interpretation efficiency without significant loss of critical diagnostic information.
2Productivity
If region selection is performed to improve interpretation efficiency, then time is saved, but risk of missing important lesions increases
Solution Approach 1:
The patent modifies the selection criteria by introducing a comprehensive evaluation value that incorporates lesion size, degree of certainty from AI detection, and anatomical importance. By adjusting these parameters and their weighting, the system can maintain high reliability in lesion detection while improving interpretation efficiency through intelligent filtering of regions.
Solution Approach 2:
The system provides feedback to users about the selection criteria and evaluation values, allowing doctors to understand why certain regions are selected or excluded. This transparency maintains trust and allows users to override selections if needed, preserving reliability while benefiting from efficiency improvements.
Data Source
AI summary
An information processing apparatus detects regions of body parts and a lesion from a medical image, derives an evaluation value for each of the regions of the body parts overlapping the lesion, based on a degree of certainty of each partial image constituting the lesion in the medical image, and selects the region of at least one body part detected from the medical image based on the evaluation value.