Endoscope Image Processing for Gastric Cancer Diagnosis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In medical endoscope systems, diagnosing gastric cancer is challenging due to minimal color differences between atrophic and normal mucosa in intermediate stages, and existing technologies fail to adequately enhance vascular visibility alongside color differences.
Innovation Solution
An image processing device and endoscope system that employs a pseudo-color display process, selective expansion processing, and specific light quantity control to emphasize color differences and improve vascular visibility by adjusting angles and radial coordinates in the feature space, and using a light source that includes violet, blue, and green lights with controlled intensities to enhance image signals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a color difference enhancing process is applied to increase the difference between signal ratios in different color ranges, then the color difference between atrophic portion and normal portion is enhanced, but vascular visibility is not sufficiently improved
Solution Approach 1:
The image processing is segmented into two distinct functional modules: a color difference enhancing process that operates on signal ratios in different color ranges to enhance atrophic portion detection, and a frequency filtering process that operates separately to enhance vascular patterns. This segmentation allows each process to optimize for its specific function without interfering with the other, thereby resolving the contradiction between color difference enhancement and vascular visibility.
Solution Approach 2:
The patent merges the color difference enhancing process and the frequency filtering process into a single integrated image processing system. The color difference enhancing process modifies signal ratios to highlight atrophic portions, while the frequency filtering process simultaneously enhances vascular patterns through frequency domain filtering. By combining these processes, the system achieves both improved color difference detection and enhanced vascular visibility.
2Measurement precision
If multiple image processing processes are applied to enhance both color difference and vascular patterns, then diagnostic accuracy is improved, but device complexity increases
Solution Approach 1:
The image processing device is designed with multi-functionality to perform both color difference enhancement and frequency filtering within a single integrated system. The processor is configured to execute multiple processing algorithms simultaneously, allowing one device to provide comprehensive diagnostic information including both atrophic portion detection and vascular pattern enhancement, thereby improving diagnostic accuracy without requiring multiple separate devices.
Solution Approach 2:
The system utilizes parameter changes in the frequency domain to achieve vascular pattern enhancement. By transforming the image signal to the frequency domain, applying selective filtering based on frequency parameters, and then transforming back, the system enhances vascular patterns through parameter manipulation rather than complex spatial processing. This approach maintains relative system simplicity while achieving enhanced diagnostic capability.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
First RGB image signals are subjected to an input process. First color information is obtained from the first RGB image signals. Second color information is obtained by a selective expansion processing in which color ranges except a first color range are moved in a feature space formed by the first color information, the first color information that represents each object in a body cavity being distributed in the each color range. The second color information is converted to second RGB signals. A red display signal, a green display signal and a blue display signal are obtained by applying a pseudo-color display process to the second RGB signals.