Capsule Endoscopy Image Processing for Lesion Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The challenge lies in efficiently processing and analyzing vast amounts of time-series images, particularly from medical devices like capsule endoscopes, where numerous images need to be filtered to highlight critical areas such as lesions or mucosal regions, while minimizing the observation of non-interesting areas like bubbles or stools, to reduce the burden on observers.
Innovation Solution
An image processing device and method that detects interest areas, calculates feature amounts, classifies these areas into groups based on their features and time-series positions, selects representative areas, and outputs these as representative images, thereby prioritizing the display of critical regions and reducing unnecessary image observation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If all time-series images are observed sequentially, then complete coverage of the examination area is achieved, but the observation time and observer burden increase significantly
Solution Approach 1:
The patent extracts and displays only representative images containing interest areas (lesions, mucosal regions) from the entire time-series image group, separating critical diagnostic information from non-critical images. This extraction principle reduces the number of images requiring observer attention while maintaining examination completeness.
Solution Approach 2:
The system performs preliminary automatic analysis of time-series images to detect and identify interest areas before observer review. By pre-processing images to locate critical regions and generate representative image sets, the system prepares optimized display content in advance, reducing subsequent observation time.
2Measurement precision
If the capturing rate is increased to improve image quality, then more detailed images are obtained, but the total number of images and data processing burden increase
Solution Approach 1:
The patent extracts only essential diagnostic information from high-volume image data by identifying and displaying representative images containing interest areas. This extraction reduces the effective data processing burden while preserving measurement precision for critical regions.
Solution Approach 2:
The system applies partial action by processing and analyzing only relevant portions of the image data - specifically focusing on detecting interest areas in representative images rather than processing every single image in detail. This selective processing maintains diagnostic quality while reducing overall computational burden.
3Reliability
If all images are displayed for review, then no diagnostic information is lost, but the ease of operation and efficiency decrease due to overwhelming volume
Solution Approach 1:
The patent extracts and presents only representative images containing diagnostic interest areas, removing non-critical images from the review set. This extraction maintains diagnostic accuracy by preserving all essential information while dramatically improving operation efficiency through reduced image volume.
Solution Approach 2:
The system applies local quality by enhancing and prioritizing display of regions containing interest areas while reducing emphasis on non-critical areas. Representative images are selected and presented with focused attention on diagnostically important regions, improving operational efficiency without sacrificing diagnostic accuracy.
Data Source
AI summary
An image processing device includes: an interest area detector that detects interest areas included in a time-series image group captured in time series; a calculation processing unit that calculates feature amounts indicative of features of the interest areas; an area classification unit that classifies the interest areas into area groups, based on the feature amounts of the interest areas and time-series positions of time-series images including the interest areas; a group feature amount calculation unit that calculates a group feature amount indicative of a feature of each of the area groups; an area selection unit that selects one or more representative areas of the interest areas belonging to the area groups, from among the area groups; and a representative image output unit that outputs one or more representative images including the representative areas in the time-series image group.