Depolarized Region Classification in Polarization OCT
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing methods for polarization-sensitive OCT images struggle to accurately classify depolarized regions in the eyes, especially when signal intensity is degraded due to cataracts or lesions, leading to incorrect discrimination between RPE and choroid regions.
Innovation Solution
An image processing apparatus and method that includes a detection unit for identifying depolarized regions, an estimation unit for curve estimation, a discrimination unit for classifying regions based on continuity with the estimated curve, and a correcting unit allowing manual correction of automatic classification results, enabling accurate classification and display of RPE, particle, and choroid regions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If automatic classification of depolarized regions is performed based on curve estimation, then processing speed is improved, but classification accuracy deteriorates when signal intensity is degraded
Solution Approach 1:
The system implements feedback by allowing operators to provide correction inputs when automatic classification results are incorrect. The correction unit receives these inputs and adjusts the classification results accordingly, creating a closed-loop system that improves accuracy while maintaining automated processing speed.
Solution Approach 2:
The correction unit acts as an intermediary between the automatic classification system and the final classification result. It mediates by accepting operator corrections and integrating them with the automated classification output, resolving the conflict between speed and accuracy.
2Measurement precision
If manual correction functions are added to allow operator input, then classification accuracy is improved, but device complexity increases
Solution Approach 1:
The correction unit is designed to be multi-functional, serving both as a receiver of operator corrections and as a processor that integrates these corrections with automated classification results. This universal component reduces overall system complexity by consolidating multiple functions into a single unit.
Solution Approach 2:
The system provides self-service capabilities by allowing operators to directly input corrections without requiring complex manual intervention protocols. The correction unit automatically processes these inputs and updates classification results, reducing the need for additional complex control mechanisms.
3Device complexity
If discrimination is based on continuity with estimated curve, then processing simplicity is improved, but reliability deteriorates in cases of signal degradation
Solution Approach 1:
The system performs preliminary classification based on curve continuity before allowing operator correction. This preliminary action maintains processing simplicity while providing a baseline that can be improved through subsequent correction inputs, ensuring reliability even when signal degradation occurs.
Solution Approach 2:
The correction unit is prepared in advance to handle cases where signal degradation causes incorrect discrimination. By having the correction mechanism ready beforehand, the system cushions against reliability issues without complicating the primary discrimination process.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables precise classification and correction of depolarized regions, improving the accuracy of RPE and choroid region identification even in cases of degraded signal intensity, facilitating better diagnosis and visualization.
Implementation Method 1
The polarization sensitive OCT divides interfering light into two linearly-polarized light beams which are orthogonal to each other and detects the linearly-polarized light beams using light modulated into circularly-polarized light
Implementation Method 2
The depolarization is seen to be caused by random changing of a direction and a phase of polarization due to reflection of measurement light in fine structures (melanin, for example) in tissues
Data Source
AI summary
An image processing apparatus includes a detection unit configured to detect a depolarized region in a polarization tomographic image of a subject's eye, an estimation unit configured to estimate a curve using the extracted depolarized region, a discrimination unit configured to discriminate the extracted depolarized region as a region including the estimated curve and a region which is discontinuous with the region including the estimated curve, and a correcting unit configured to correct at least a portion of a result of the discrimination representing the discontinuous region to a result of discrimination representing another region.