Fundus Image Processing for Choroidal Vessel Discrimination
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing methods struggle to effectively analyze and visualize the choroidal vasculature, which is crucial for understanding eye health, as they often fail to distinguish between retinal and choroidal blood vessels.
Innovation Solution
An image processing method that extracts line-shaped and lump-shaped portions of choroidal vasculature from fundus images, integrating them to generate a vascular image that clearly depicts choroidal blood vessels, using techniques like de-noising, black hat filtering, in-painting, and contrast-limited adaptive histogram equalization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing image processing methods are used to visualize blood vessels in fundus images, then general vascular structures can be seen, but choroidal vasculature cannot be selectively distinguished from retinal blood vessels
Solution Approach 1:
The patent segments the vascular structures into two distinct categories: line-shaped portions (retinal blood vessels) and lump-shaped portions (choroidal vasculature including vortex veins). By applying different extraction algorithms tailored to each shape characteristic, the method achieves selective visualization of choroidal vasculature while separating it from retinal blood vessels, thereby resolving the contradiction between general visibility and selective discrimination.
2Measurement precision
If multiple image processing steps are applied to extract choroidal vasculature, then visualization accuracy improves, but processing complexity increases
Solution Approach 1:
The processing system is segmented into distinct functional modules: a line-shaped portion extraction section for retinal vessels, a lump-shaped portion extraction section for choroidal vasculature, and a combining section. Each module performs a specific task with dedicated algorithms, which simplifies the overall system architecture while maintaining high extraction accuracy through specialized processing for each vascular type.
Solution Approach 2:
The patent extracts specific features (line-shaped and lump-shaped portions) from the fundus image using targeted extraction algorithms. By taking out only the relevant vascular components needed for choroidal analysis and discarding redundant information, the system achieves high precision without requiring overly complex processing of the entire image.
Data Source
AI summary
An image processing method including acquiring a fundus image, extracting a first area including a first feature from the fundus image, extracting a second area including a second feature different from the first feature from the fundus image, and generating a combined image in which the extracted first area and the extracted second area are combined.