Endoscope Correction Data Generation for Lesion Score Precision
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing electronic endoscope systems exhibit significant variations in lesion index scores when imaging the same lesion site using different systems, due to individual differences in optical components, leading to degradation in score calculation precision.
Innovation Solution
A correction data generation method and apparatus that acquires and plots image data points in a predetermined color space, calculates a correction value based on the distance between data points and target points, and stores this value to correct pixel values in captured images, thereby reducing score variations across different endoscope systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If white balance adjustment is performed to improve color reproduction, then color accuracy is improved, but individual differences in optical components cause remaining errors that degrade score calculation precision
Solution Approach 1:
The patent performs preliminary color space conversion to a predetermined color space (e.g., L*a*b* or HSV) before white balance adjustment and score calculation. This preliminary action transforms the data into a space where optical component variations have less impact, allowing subsequent processing to achieve both color accuracy and reliable scoring.
Solution Approach 2:
The patent changes the color representation parameters by converting from RGB to a predetermined color space that separates luminance and chrominance components. This parameter transformation allows the system to adjust white balance while maintaining stable color characteristics that are less sensitive to optical component variations, thereby improving both color accuracy and score reliability.
2Adaptability or versatility
If different electronic endoscope systems are used to image the same lesion site, then imaging flexibility is improved, but large variations occur in lesion index score values
Solution Approach 1:
The patent creates a universal color space representation that can be applied across different endoscope systems. By converting all systems' data to a predetermined color space with standardized white balance processing, the system achieves multi-functionality where the same scoring algorithm can reliably process images from various endoscope models, maintaining score consistency while preserving imaging flexibility.
Solution Approach 2:
The patent standardizes color parameters across different systems by converting to a predetermined color space and applying unified white balance adjustment. This parameter standardization ensures that the same lesion site imaged by different endoscope systems produces consistent score values, while still allowing each system to maintain its imaging capabilities.
3Difficulty of detecting and measuring
If tone enhancement processing is applied to widen dynamic range near lesion boundaries, then lesion detectability is improved, but complexity of color space conversion and processing increases
Solution Approach 1:
The patent segments the image processing into distinct stages: first converting to a predetermined color space, then applying tone enhancement to specific channels or regions. This segmentation allows tone enhancement to be applied selectively to widen the dynamic range near lesion boundaries without unnecessarily complicating the entire processing pipeline, maintaining lesion detectability while managing complexity through structured processing steps.
Data Source
AI summary
Provided is a correction data generation method that includes acquiring captured image data by imaging an indicator that is related to a predetermined illness; plotting a real imaging data point that corresponds to the acquired captured image data in a predetermined color space that is associated with the predetermined illness in accordance with a color component of the data point; calculating a correction value for correcting the values of pixels that make up a captured image captured by an electronic endoscope based on the distance between the data point and a predetermined target point in the predetermined color space; and storing the calculated correction value.