Ultra-Wide Fundus Image Segmentation for Abnormality Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies struggle to effectively detect abnormal areas in fundus images, particularly in ultra-wide field images, which are crucial for diagnosing eye conditions.
Innovation Solution
The method involves dividing the fundus region of an ultra-wide field image into multiple areas using blood vessel boundaries or choroidal vascular networks, generating attribute information for each area, and instructing display modes based on these attributes to highlight normal and abnormal regions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the fundus region is divided into multiple areas using watershed segmentation, then the detection precision of abnormal areas is improved, but the device complexity increases
Solution Approach 1:
The fundus region is divided into multiple areas using watershed segmentation based on choroidal vascular networks. This segmentation allows for region-specific analysis and display mode assignment, improving detection precision by treating different fundus areas independently with appropriate display modes for abnormality detection.
Solution Approach 2:
Different display modes are assigned to different divided areas based on their attributes. Normal areas, abnormal areas, and areas requiring attention each receive specific display modes (e.g., original image mode, inverted mode, highlighted mode), allowing local optimization of detection precision for each region type.
2Ease of operation
If different display modes are assigned to different areas, then the ease of operation is improved, but the device complexity increases
Solution Approach 1:
The system automatically assigns appropriate display modes to different fundus areas based on their attributes (normal, abnormal, attention-required). This local differentiation improves ease of operation by presenting information in the most suitable format for each region, reducing the cognitive load on operators.
Solution Approach 2:
The system performs automatic area division and mode assignment based on image analysis, eliminating the need for manual configuration. The processor automatically identifies areas and assigns display modes, making the system self-sufficient and improving ease of operation without requiring user expertise in image analysis.
3Loss of information
If attribute information is generated for each area, then the loss of information is reduced, but the processing time increases
Solution Approach 1:
By dividing the fundus region into multiple areas, the system can generate attribute information for each area independently and efficiently. This segmentation allows parallel processing of different regions, reducing overall processing time while preserving detailed attribute information for each area, thus minimizing information loss.
Solution Approach 2:
The system performs preliminary area division and attribute generation before final display and analysis. By pre-processing the image into divided areas with assigned attributes and display modes, the system prepares the data structure in advance, enabling faster subsequent analysis and reducing the time required for detailed examination.
Data Source
AI summary
A processor divides a fundus region of an ultra-wide field fundus image into plural areas including at least a first area and a second area, generates first attribute information indicating an attribute of the first area and second attribute information indicating an attribute of the second area, and generates first mode instruction information to instruct display of the first area in a first mode corresponding to the first attribute information, and generates second mode instruction information to instruct display of the second area in a second mode corresponding to the second attribute information.