Channel-Coded OCT Slabs for Accurate GA Segmentation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for detecting geographic atrophy (GA) in optical coherence tomography (OCT) data are inaccurate and time-consuming due to reliance on single-factor analysis, such as sub-RPE reflectivity, which is prone to errors from factors like choroidal blood vessels and retinal opacities, requiring subsequent B-scan reviews.

Innovation Solution

A method that integrates multiple aspects of OCT volumetric information into different image channels, combining metrics like RPE integrity, retinal thinning, and optical attenuation coefficient to create a channel-coded image for more accurate GA detection and segmentation, using machine learning models like U-Net for automated analysis.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If single-factor analysis (sub-RPE reflectivity) is used for GA detection, then the method is simple and fast, but accuracy is poor due to errors from choroidal blood vessels and retinal opacities

Engineering Contradiction:
ImproveGA detection accuracyVSAvoidanalysis method complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the OCT volumetric data into multiple distinct metrics (sub-RPE reflectivity, inner RPE reflectivity, retinal thickness, optical attenuation coefficient) and analyzes each separately before integrating them. This segmentation allows the system to handle complex information systematically, improving GA detection accuracy by evaluating multiple factors rather than relying on a single potentially misleading metric.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent merges multiple separate OCT metrics into a unified channel-coded image where each metric is represented in a different image channel. This combining approach allows the system to integrate information from sub-RPE reflectivity, inner RPE reflectivity, retinal thickness, and optical attenuation coefficient simultaneously, creating a comprehensive GA detection method that overcomes the limitations of single-factor analysis.

Inventive Principle:
Principle #5Merging (Combining)

2Productivity

If manual segmentation of GA regions is performed, then segmentation accuracy is high, but the task is time-consuming

Engineering Contradiction:
Improvesegmentation speedVSAvoidsegmentation accuracy
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The patent creates a synthesized channel-coded image that copies and integrates information from multiple OCT metrics into a single image representation. This copied and integrated data structure enables automated algorithms to process GA segmentation efficiently, achieving both high speed and maintained accuracy by working with the pre-processed multi-metric image rather than raw volumetric data.

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The patent replaces the mechanical manual segmentation process with an automated machine learning-based segmentation system. The channel-coded image serves as input for automated algorithms that can rapidly and accurately identify GA regions without human intervention, substituting the time-consuming manual process with computational methods that maintain or improve segmentation precision.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

3Reliability

If B-scan by B-scan review is performed to confirm GA presence, then diagnostic accuracy is improved, but time required increases significantly

Engineering Contradiction:
ImproveGA diagnosis reliabilityVSAvoidtime for B-scan review
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent performs preliminary analysis by pre-processing the OCT volumetric data into a channel-coded image that integrates multiple metrics before the actual GA detection process. This preliminary action organizes and optimizes the data structure, enabling rapid and reliable GA detection without requiring time-consuming B-scan by B-scan review, as the integrated image already contains the necessary multi-factor information.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The channel-coded image serves multiple functions simultaneously: it represents sub-RPE reflectivity, inner RPE reflectivity, retinal thickness, and optical attenuation coefficient all in one image structure. This multi-functionality allows the system to perform comprehensive GA analysis in a single processing step, eliminating the need for separate B-scan reviews while maintaining high diagnostic reliability.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP4143781B1Oct en face pathology segmentation using channel-coded slabs
Publication Date: 2025.12.10 CARL ZEISS MEDITEC INC
  • EP4143781B1 patent drawingFigure 1~4B
  • EP4143781B1 patent drawingFigure 3
  • EP4143781B1 patent drawingFigure 5

AI summary

A system and method for use with optical coherence tomography (OCT) data to identify a target pathology extracts multiple pathology-characteristic images from the OCT data. The extracted pathology-characteristic images may include a mixture of OCT structural images (including retinal layer thickness information) and OCT angiography images. Optionally, other pathology-characteristic images and data maps (mapped to corresponding positions in the OCT data), such as fundus images and visual field test maps may be accessed as additional pathology-characteristic images. Each pathology-characteristic image defines a different image channel (e.g., "color channel") per pixel in a composite, channel-coded image, which is then used to train a neural network to search for the target pathology in OCT data. The trained neural network may then receive new composite, channel-coded image and identify/segment the target pathology within the new channel-coded image.