Endoscope Image Processing for Lesion Detection Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting lesion sites during endoscopic examinations often fail to accurately identify areas of interest, particularly those that are difficult to distinguish from single images.
Innovation Solution
An image processing device that acquires and analyzes time-series images from an endoscope, calculates a score for the likelihood of a point of interest using a likelihood ratio calculation model, and classifies images based on this score to enhance detection accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If individual images are analyzed one by one to detect lesion sites, then the analysis process is simple and fast, but the detection accuracy is insufficient for lesions difficult to distinguish from single images
Solution Approach 1:
The patent merges multiple captured images in the time series to perform classification. By combining information from multiple images rather than analyzing them individually, the system achieves higher detection accuracy for lesion sites that are difficult to distinguish from single images, while managing complexity through efficient processing methods.
2Measurement precision
If multiple captured images in time series are used for classification, then the detection accuracy of lesion sites is improved, but the processing time and computational load increase
Solution Approach 1:
The patent performs preliminary classification of captured images based on scores calculated from multiple images in the time series. By pre-processing and scoring images before final classification, the system reduces the computational burden during real-time processing while maintaining high detection accuracy for lesion sites.
3Reliability
If multiple captured images are classified based on calculated scores, then the reliability of lesion detection is enhanced, but the device complexity increases
Solution Approach 1:
The patent changes the parameter used for classification from simple binary detection to score-based classification. By calculating scores from multiple captured images and using these scores for classification, the system enhances detection reliability while managing complexity through efficient score calculation and classification algorithms.
Data Source
AI summary
An image processing device 1X includes an acquisition means 30X, a score calculation means 31X, and a classification means 34X. The acquisition means 30X acquires each captured image which captures an examination target by an imaging unit provided on an endoscope. The score calculation means 31X calculates a score concerning a likelihood of an existence of a point of interest in captured images based on a plurality of captured images which are in a time series. The classification means 34X classifies the plurality of captured images based on the score.