Conjunctival Hyperemia Detection via Blood Vessel Node Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing devices for determining conjunctival hyperemia lack accuracy in assessing the severity and localization of hyperemia, often relying solely on red pixel values or total eye area, which can lead to unreliable results and fail to provide detailed analysis by parts of the eye.
Innovation Solution
An electronic device equipped with a camera and processor that identifies and analyzes blood vessels by dividing them into nodes based on cross points, calculates their sizes, and determines hyperemia degrees by comparing these sizes to a predetermined unit area, considering locations and thickness, to provide a detailed and trustworthy assessment of conjunctival hyperemia.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If the hyperemia degree is calculated based on red pixel values in the eye image, then the measurement process is simple, but the accuracy of hyperemia detection deteriorates when hyperemia is not severe
Solution Approach 1:
The patent segments the conjunctiva into multiple regions (upper, lower, nasal, temporal) and further divides blood vessels into nodes based on cross points. This segmentation allows precise localization and measurement of hyperemia in specific areas rather than treating the entire eye as a single unit, thereby improving detection accuracy while maintaining reasonable complexity.
Solution Approach 2:
The patent applies local quality analysis by measuring hyperemia degree separately for different regions of the conjunctiva (upper, lower, nasal, temporal) and different sizes of blood vessels. This enables differentiated assessment of hyperemia in various locations, improving overall measurement precision without requiring excessively complex global analysis.
2Device complexity
If the hyperemia degree is calculated based on total area of the eye, then the calculation is straightforward, but the ability to provide hyperemia degree by parts of the eye deteriorates
Solution Approach 1:
The patent divides the conjunctiva into four distinct regions (upper, lower, nasal, temporal) and further segments blood vessels into nodes based on cross points. This multi-level segmentation preserves detailed localization information about where hyperemia occurs in the eye, preventing information loss while keeping the calculation framework organized and manageable.
Solution Approach 2:
The patent transitions from a single global hyperemia measurement to a multi-dimensional assessment by analyzing hyperemia in four different conjunctival regions and categorizing blood vessels by size. This dimensional expansion provides comprehensive localization information without overwhelming computational complexity.
3Device complexity
If only a result value of hyperemia analysis is provided to the user, then the output is simple, but the user's trust in the result value deteriorates
Solution Approach 1:
The patent implements feedback by providing users with detailed information about the hyperemia analysis process, including the detected blood vessel nodes, their locations in different conjunctival regions, and the calculated hyperemia degree. This transparent feedback mechanism allows users to verify the analysis logic and build trust in the results while maintaining relatively simple output presentation.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Disclosed are an electronic device and a method for determining a degree of conjunctival hyperemia using the same. The electronic device includes a camera and a processor configured to acquire an image including an eye captured by the camera, identify one or more blood vessels included in the image, and determine a degree of conjunctival hyperemia based on sizes of the identified one or more blood vessels.