Adaptive Polyp Detection in Endoscopic Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current medical image processing methods for detecting lesions in endoscopic images face challenges in accurately identifying polyps due to variations in color tone changes and three-dimensional shape features, often requiring rigid detection standards that may lead to oversight or false positives.
Innovation Solution
A medical image processing device and method that incorporate a lesion candidate region detection based on color signals, incidental region detection for attributes like color tone changes, and adaptive detection standards, adjusting thresholds based on the presence of color tone changes and three-dimensional shape information to enhance polyp detection accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If rigid detection standards are used for polyp detection, then the detection process is simple and fast, but detection accuracy decreases due to false positives and oversight
Solution Approach 1:
The patent implements dynamic detection standards that automatically adjust based on the detected characteristics of the medical image. The system evaluates color tone changes and three-dimensional shape features to determine whether to apply a first detection standard (when color tone changes are detected) or a second detection standard (when only shape features are present). This dynamic adaptation resolves the contradiction by maintaining high accuracy through conditional standard selection while avoiding the complexity of manually designing multiple fixed detection protocols.
Solution Approach 2:
The system changes detection parameters (detection standards) based on the presence or absence of specific features in the medical image. When color tone changes are detected, the system switches to a first detection standard with different threshold criteria compared to the second standard used when only shape features are present. This parameter change approach allows the system to optimize detection accuracy for different polyp types without requiring a completely complex detection framework.
2Reliability
If color information is utilized for lesion detection, then detection reliability improves, but processing time increases due to additional analysis requirements
Solution Approach 1:
The patent segments the detection process into distinct stages: first detecting color tone changes and three-dimensional shape features separately, then using these results to determine which detection standard to apply. By segmenting the analysis and only performing comprehensive color-based detection when necessary (when color tone changes are detected), the system improves reliability for color-related lesions while minimizing unnecessary processing time for lesions that can be detected through shape features alone.
Solution Approach 2:
The system applies partial action by using color information only when it is likely to be relevant (when color tone changes are detected in the image). When only shape features are present, the system uses the simpler second detection standard without performing extensive color analysis. This partial application of color-based detection maintains high reliability for color-related polyps while avoiding the time cost of comprehensive color analysis for all images.
3Measurement precision
If detection standards are adjusted for different lesion types, then false positives are reduced, but the detection system becomes more complex
Solution Approach 1:
The system uses feedback from the initial detection of color tone changes and shape features to determine which detection standard to apply. The detection process continuously evaluates the image characteristics and adjusts the detection criteria accordingly. This feedback mechanism allows the system to reduce false positives by selecting appropriate standards for different lesion types while keeping the overall system structure relatively simple through automated decision-making rather than complex manual configuration.
Data Source
AI summary
An image processing device includes; a lesion candidate region detecting section that detects a lesion candidate region based on at least one color signal in a medical image including a plurality of color signals; an incidental region detecting section that detects an incidental region that arises because of incidental attributes accompanying a lesion from the medical image; and a detection standard changing section that changes a detection standard when detecting a lesion from the lesion candidate region in accordance with a detection result of the incidental region.