3D Fundus Image Mapping for Peripheral Lesion Positioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing systems struggle with accurately positioning fundus structures and pathological lesions at the peripheral portion of the eye in ultra-wide field (UWF) fundus images due to distortion, making it difficult to ascertain their location on the eyeball.
Innovation Solution
An image processing method that includes acquiring a two-dimensional fundus image, aligning it with an eyeball model, and creating data to represent the movement of points from the fundus image to the eyeball model, using reverse stereographic projection and predefined transformation equations to correct distortions and generate 3D video data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If reverse stereographic projection is used to map 2D fundus image points to 3D eyeball model points, then positioning accuracy of fundus structures is improved, but computational complexity increases
Solution Approach 1:
The patent pre-calculates and stores transformation equations that map 2D fundus image coordinates to 3D eyeball model coordinates. These pre-computed transformation relationships are then applied during image processing, avoiding the need for complex real-time 3D geometric calculations while maintaining high positioning accuracy for fundus structures and lesions.
Solution Approach 2:
The patent creates a 3D eyeball model as a virtual copy of the actual eyeball geometry, and maps 2D fundus image points to corresponding points on this virtual 3D model. This allows accurate positioning without directly manipulating complex 3D spatial coordinates, as the transformation follows predefined geometric relationships between the 2D image plane and 3D spherical model.
2Measurement precision
If 3D video data is generated from 2D fundus images, then visualization accuracy is improved, but processing time increases
Solution Approach 1:
The patent pre-establishes the geometric transformation relationships and mapping algorithms between 2D fundus images and 3D eyeball models before actual processing. This allows the 3D visualization to be generated more efficiently during clinical use, as the computationally intensive aspects of the transformation have already been prepared in advance.
Data Source
AI summary
Image processing performed by a processor and including acquiring a two-dimensional fundus image, acquiring a second point on an eyeball model corresponding to at least one first point of the two-dimensional fundus image, and creating data to represent a process to move the first point to the second point.