Whole Slide Image Creation Using Pattern Matching Magnification Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Whole slide image creation devices take a long time to specify and capture regions of interest for pathology diagnosis due to the need to observe specimens at varying magnifications and determine optimal magnification for imaging.
Innovation Solution
A device with a first image-capture controller, storage, calculator, and magnification selector that captures images at multiple magnifications, performs pattern matching to select the optimal magnification, and captures the whole image in segmented regions along a spiral line, using teacher images to efficiently determine the appropriate magnification for pathology specimen imaging.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the device captures images at multiple magnifications and performs pattern matching to determine optimal magnification, then the quality and accuracy of pathology diagnosis is improved, but the time required to create whole slide images increases significantly
Solution Approach 1:
The system performs preliminary actions by capturing a low-magnification overview image first, then uses pattern matching with teacher images to predict the optimal magnification before actually capturing high-magnification images. This preliminary determination of capture magnification based on low-magnification image analysis avoids the time-consuming process of capturing and evaluating multiple magnifications sequentially
Solution Approach 2:
The system creates virtual copies through pattern matching between low-magnification images and stored teacher images at various magnifications. By matching patterns from the overview image against pre-stored teacher images, the system determines which magnification would yield the best diagnostic results without physically capturing images at all magnifications, thus reducing time while maintaining diagnostic accuracy
2Manufacturing precision
If the device captures the entire pathology specimen at high magnification, then the detail quality for diagnosis is improved, but the time and resources required increase significantly
Solution Approach 1:
The system segments the pathology specimen into multiple regions and processes them differently. The overview image is divided into grid sections that are matched against teacher images to determine which regions require high-magnification capture. This segmentation allows the system to apply high-detail capture only to relevant regions rather than the entire specimen, improving efficiency while maintaining detail quality where needed
Solution Approach 2:
The system applies different quality levels to different regions of the specimen based on local characteristics. Regions identified through pattern matching as containing diagnostically important features are captured at high magnification with detailed quality, while other regions are handled at lower magnification. This local quality approach ensures diagnostic detail is preserved where necessary while improving overall productivity
Data Source
AI summary
A whole slide image creation device including a first image-capture control unit for capturing an image-capture region for a pathological specimen at a first magnification and a second magnification; a storage unit for storing a first training image corresponding to the first magnification and a second training image corresponding to the second magnification; a calculation unit for pattern matching between a first image captured at the first magnification and the first training image to calculate a first degree-of-match, and between a second image captured at the second magnification and the second training image to calculate a second degree-of-match; a magnification selection unit for selecting the first magnification or the second magnification for capturing the entire pathological specimen on the basis of the first degree-of-match and the second degree-of-match; and a second image-capture control unit for capturing the entire pathological specimen at the selected magnification for each segmented region.