Optical Coherence Tomography Layer Segmentation Under Weak Contrast

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current OCT segmentation methods struggle with accurate segmentation of layer boundaries, particularly for edges with weak contrast, such as the boundary between the posterior lens (capsule) and the vitreous, due to speckles and complicated pathologies like dense cataracts, leading to reduced accuracy or impossibility of segmentation.

Innovation Solution

A method and apparatus that utilize feature integration to minimize noise features, enhancing the true edge by generating first, second, and third feature images through gradient calculations and mathematical operations, improving segmentation performance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional segmentation methods are used on OCT images, then the processing is simple and fast, but the segmentation accuracy is reduced or impossible for cases with speckles and dense cataracts

Engineering Contradiction:
Improvesegmentation accuracyVSAvoidsegmentation method complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent applies segmentation by dividing the OCT image processing into multiple stages: generating first features from the original image, generating second features from integrated image data, and combining them into third features for final segmentation. This multi-stage segmentation approach improves accuracy for weak contrast edges while managing complexity through systematic decomposition of the segmentation task.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent transitions from processing two-dimensional OCT images to three-dimensional volumetric data by integrating image data across multiple slices. This dimensional expansion allows the system to leverage information from adjacent slices to enhance edge detection and segmentation accuracy, particularly for boundaries that are indistinct in single 2D slices.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Measurement precision

If manual segmentation is performed, then segmentation accuracy can be maintained, but the process is time consuming and subjective

Engineering Contradiction:
Improvesegmentation accuracyVSAvoidsegmentation time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent implements self-service through automated algorithms that perform segmentation without manual intervention. The system automatically generates multiple feature types, integrates them, and produces segmentation results autonomously. This eliminates the time-consuming and subjective nature of manual segmentation while maintaining consistent accuracy across different cases.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent incorporates feedback mechanisms where segmentation results are evaluated and used to refine the segmentation process. The system iteratively adjusts parameters and re-processes images to improve accuracy, mimicking the adaptive nature of manual segmentation while maintaining automated efficiency.

Inventive Principle:
Principle #23Feedback

3Measurement precision

If feature integration is performed to enhance weak contrast edges, then segmentation accuracy improves, but the processing complexity and computational load increase

Engineering Contradiction:
Improveedge detection accuracyVSAvoidcomputational energy
Core Design Contradiction:
Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The patent applies partial action by selectively generating different types of features only where needed in the image. Rather than processing the entire volumetric dataset with all feature generation algorithms uniformly, the system applies processing selectively to regions containing weak contrast edges or potential segmentation challenges, reducing overall computational energy while maintaining accuracy where it matters most.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP3716835B1Improving segmentation in optical coherence tomography imaging
Publication Date: 2025.09.24 ALCON INC
  • EP3716835B1 patent drawingFigure 1
  • EP3716835B1 patent drawingFigure 2
  • EP3716835B1 patent drawingFigure 3

AI summary

A method for improving segmentation in optical coherence tomography imaging. The method comprises obtaining an OCT image of imaged tissue, generating a first feature image for at least a portion of the OCT image, and generating a second feature image for at least the portion of the OCT image, based on either the OCT image or the first feature image, by integrating image data in a first direction across the OCT image or first feature image. A third feature image is generated as a mathematical function of the first and second feature images, and layer segmentation for the OCT image is performed, based on the third feature image.