In-Vivo Image Stream Filtering for Gastrointestinal Content Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In-vivo imaging systems face challenges in efficiently detecting and filtering out intestinal content from image streams captured during gastrointestinal tract imaging, leading to obscured tissue visualization and increased viewing time for healthcare professionals.
Innovation Solution
A method and system that calculate pixel and image content scores to select and display images with high probability of non-obscured regions, using a processing unit to analyze image streams from ingestible capsules, and a classifier trained on marked images to differentiate between content and tissue pixels, thereby reducing the number of images displayed and focusing on clinically significant frames.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If all images from the image stream are displayed to healthcare professionals, then complete visual information is provided, but viewing time increases and diagnostic efficiency decreases
Solution Approach 1:
The system extracts and removes images that are heavily obscured by intestinal content from the image stream, retaining only images with sufficient visible tissue for diagnostic evaluation. This extraction process reduces the total number of images while preserving clinically relevant information.
Solution Approach 2:
The content detection algorithm performs preliminary analysis of each image to assess the degree of obscuration by intestinal content before images are presented to healthcare professionals. This preliminary filtering action prevents unnecessary viewing of obscured images, saving time while maintaining information quality.
2Reliability
If images obscured by intestinal content are included in the display, then complete coverage of the GI tract is maintained, but diagnostic accuracy decreases due to poor tissue visualization
Solution Approach 1:
The system extracts and excludes images where intestinal content obscures more than a threshold percentage of the tissue area. By removing these low-quality images, the system maintains diagnostic accuracy by ensuring only images with sufficient tissue visualization are presented for interpretation.
Solution Approach 2:
The system changes the parameter of image quality by filtering based on the proportion of visible tissue versus obscured content. Images are retained only if they meet a minimum threshold of tissue visibility, transforming the image stream from including all captured frames to including only diagnostically useful frames.
3Loss of time
If the image stream is filtered to remove content-obscured images, then viewing time is reduced, but the complexity of image processing increases
Solution Approach 1:
The system replaces manual review of all images by healthcare professionals with an automated content detection algorithm. This substitution uses computational methods to assess image quality and filter obscured images, reducing the mechanical burden on human reviewers while managing processing complexity through algorithmic approaches.
Solution Approach 2:
The content detection system performs self-service by automatically analyzing and filtering images without requiring human intervention for each individual image assessment. The algorithm independently evaluates each image for obscuration levels and makes filtering decisions, reducing overall processing complexity compared to manual review methods.
Data Source
AI summary
A system and method for detecting in-vivo content includes an in-vivo imaging device for capturing a stream of image frames in a GI tract, a content detector for detecting and/or identifying one or more image frames from the stream of image streams that may show substantially only content, a display selector to remove detected frames from the image stream, and a monitor to display the remaining image frames as a reduced image stream.