Choroidal Vascular Image Analysis for Fundus Watershed Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing technologies lack effective methods for analyzing choroidal blood vessels from a fundus image, particularly in identifying key vascular structures like the watershed of the choroidal vascular network.

Innovation Solution

An image processing method and device that acquire a fundus image, generate a choroidal vascular image, and detect the watershed of the choroidal vascular network using various detection methods, including identifying key landmarks like the macula and optic nerve head, and analyzing blood vessel density and direction to identify watersheds.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If conventional image processing methods are used on fundus images, then general image analysis is possible, but choroidal vascular network analysis cannot be performed

Engineering Contradiction:

Improvecapability to analyze choroidal blood vesselsVSAvoiddetection accuracy of watershed and vortex veins

Core Design Contradiction:

Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The patent segments the fundus image processing into distinct stages: first generating a choroidal vascular image by removing retinal vascular components, then detecting watersheds in the choroidal network, and finally identifying vortex veins. This segmentation allows each processing stage to be optimized independently, enabling choroidal analysis while maintaining high detection precision through specialized algorithms for each task.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces an intermediary choroidal vascular image as a intermediate representation between the original fundus image and the final watershed/vortex vein detection results. This intermediary image isolates choroidal vessels by removing retinal vascular components, serving as a specialized mediator that enables precise choroidal analysis without interfering with subsequent detection tasks.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Measurement precision

If multiple detection methods are implemented for watershed detection, then detection accuracy improves, but processing complexity increases

Engineering Contradiction:

Improvewatershed detection accuracyVSAvoidprocessing system complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent implements multiple watershed detection methods (first detection method using density gradient, second detection method using skeletonization) that can be dynamically selected or combined based on image characteristics. This dynamic approach allows the system to adapt to different choroidal vascular patterns, improving detection accuracy while managing complexity through conditional execution rather than always processing all methods.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The patent employs different detection algorithms with varying parameters - the first detection method uses density gradient calculations with specific threshold parameters, while the second method uses skeletonization with different structural parameters. By changing detection parameters based on image characteristics, the system achieves high accuracy across diverse cases without requiring a single overly complex unified algorithm.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS20250344948A1Image processing method, image processing device, and program

Publication Date: 2025.11.13 NIKON CORP

AI summary

A processor acquires a fundus image, generates a choroidal vascular image from the fundus image, and detects a watershed of a choroidal vascular network in the choroidal vascular image.