Eye Image Registration Using Transition Edges Under Pupil Dilation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing ophthalmic surgical systems face challenges in accurately aligning eye images due to differences in eye orientation and pupil dilation between diagnostic and treatment stages, which can distort iris features and hinder registration accuracy.

Innovation Solution

The method employs transition-edge based registration using the outer edge of the pupil and inner edge of the sclera to align eye images, complemented by iris registration when applicable, to ensure accurate alignment despite changes in pupil dilation and lighting conditions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If iris-based registration is used to align eye images, then registration accuracy can be achieved under stable conditions, but accuracy deteriorates when pupil dilation changes or lighting conditions vary

Engineering Contradiction:

Improveregistration accuracyVSAvoidrobustness to pupil dilation and lighting changes

Core Design Contradiction:

Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The patent introduces transition edges (pupil-sclera boundary) as an intermediary reference feature that mediates between the pupil center and scleral landmarks. These transition edges serve as stable intermediaries that maintain consistent geometric relationships across different imaging conditions, enabling accurate registration without direct reliance on iris features that are sensitive to dilation and lighting variations.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent segments the eye image into distinct functional zones: pupil region, transition edge region, and scleral region. By focusing registration on the transition edge boundaries rather than the entire iris, the system divides the complex registration problem into manageable segments that are more robust to physiological changes and environmental variations.

Inventive Principle:

Principle #1Segmentation

2Adaptability or versatility

If eye images are captured under different lighting conditions and pupil dilation states, then comprehensive diagnostic and treatment imaging is achieved, but alignment accuracy between images deteriorates

Engineering Contradiction:

Improveimaging flexibilityVSAvoidalignment accuracy

Core Design Contradiction:

Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The patent identifies and utilizes the geometric parameters of transition edges (position, orientation, curvature) as invariant markers that maintain consistent relationships despite changes in pupil dilation state and lighting conditions. By basing registration on these parameter-stable transition edges rather than pixel-intensity-based iris features, the system achieves accurate alignment across variable imaging conditions.

Inventive Principle:

Principle #35Parameter changes

3Device complexity

If traditional feature-based registration is used, then simple implementation is achieved, but registration reliability deteriorates under varying eye orientation and dilation

Engineering Contradiction:

Improveimplementation simplicityVSAvoidregistration reliability

Core Design Contradiction:

Device complexityVSReliability

Solution Approach 1:

The patent replaces complex multi-feature matching algorithms with a simplified geometric approach based on transition edge detection. Instead of using numerous iris landmarks and complex correspondence algorithms, the system substitutes a more reliable but computationally simpler method that leverages the natural geometric boundaries between eye structures, achieving both simplicity and reliability.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS20260069143A1Transition edge based eye registration

Publication Date: 2026.03.12 ALCON INC

AI summary

The present disclosure relates to operations that include identifying a first transition-edge portion of a first image of an eye that corresponds to a transition edge of the eye. The operations may also include determining, based on first-portion pixel data corresponding to the first transition-edge portion, a first transition-edge representation of the transition edge. The operations may include identifying a second transition-edge portion of a second image of the eye that corresponds to the transition edge. In addition, the operations may include determining, based on second-portion pixel data corresponding to the second transition-edge portion, a second transition-edge representation of the transition edge. Moreover, the operations may include determining an alignment registration between the eye as depicted in the first image and the second image based on a transition-edge based registration that is based on a comparison between the first transition-edge representation and the second transition-edge representation.