OCT Image-Pair Filtering for Reliable Optoretinography
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional ORG processing techniques yield unreliable tissue velocity indications due to unreliable OCT image pairs, which can adversely affect the quality of ORG data and other final processing results.
Innovation Solution
A computer-implemented method and apparatus process OCT images using image quality and similarity metrics to filter out unreliable tissue velocity indications, ensuring only reliable data is used for further processing by calculating and comparing comparison values with a threshold.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional ORG processing techniques are used to process OCT images, then processing speed is maintained, but the reliability of tissue velocity indications deteriorates due to unreliable OCT image pairs
Solution Approach 1:
The patent applies preliminary action by calculating image quality metrics and similarity metrics before performing phase-based velocity calculations. This allows unreliable OCT image pairs to be identified and excluded in advance, preventing them from corrupting the final velocity measurements. The comparison value is computed beforehand to determine whether to proceed with phase processing, ensuring reliability without requiring complex post-processing corrections.
Solution Approach 2:
The patent introduces intermediary metrics (image quality metric and similarity metric) that act as mediators between the raw OCT images and the final velocity calculations. These metrics serve as a filtering layer that assesses the suitability of image pairs before they are used for phase-based measurements, thereby improving reliability without directly modifying the core velocity calculation algorithm.
2Measurement precision
If all OCT image pairs are processed to generate tissue velocity indications, then productivity is maintained, but measurement precision deteriorates due to inclusion of unreliable data
Solution Approach 1:
The patent applies the taking out principle by extracting and removing unreliable OCT image pairs from the processing pipeline based on comparison values derived from image quality and similarity metrics. By selectively excluding image pairs that do not meet reliability thresholds, the precision of the remaining velocity measurements is improved without requiring reprocessing of the entire dataset, thus maintaining acceptable productivity.
Solution Approach 2:
The patent changes the parameter of image pair selection from inclusive (processing all pairs) to selective (processing only pairs meeting quality thresholds). This is achieved by introducing comparison values based on image quality metrics and similarity metrics, which transform the processing criterion from quantity-based to quality-based, thereby improving measurement precision while managing productivity through intelligent filtering.
3Reliability
If image quality and similarity metrics are calculated to filter unreliable data, then reliability of ORG data improves, but use of energy increases due to additional processing steps
Solution Approach 1:
The patent applies partial action by calculating image quality metrics and similarity metrics only for the purpose of determining whether to proceed with phase-based velocity calculations. Rather than performing exhaustive analysis on all image pairs, the system performs just enough additional processing (calculating comparison values) to make a go/no-go decision, thereby improving reliability with minimal additional energy expenditure.
Solution Approach 2:
The patent enables the OCT processing system to self-evaluate the quality of its input data through automated calculation of image quality metrics and similarity metrics. This self-service mechanism allows the system to autonomously identify and exclude unreliable image pairs without external intervention, improving reliability while keeping the additional energy cost contained within the existing processing pipeline.
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
Improves the reliability and accuracy of ORG data by removing unreliable velocity profiles, enhancing the precision of ORG data and other processing results.
Implementation Method 1
Optical coherence tomography (OCT) is an imaging technique based on low-coherence interferometry
Implementation Method 2
a spectral interferogram resulting from an interference between light in the reference arm and light in the sample arm of the interferometer at each A-scan location is Fourier transformed to simultaneously acquire all points along the depth of the A-scan
Implementation Method 3
OCT imaging systems can also be classified as being phase-resolved, where both the intensity and phase of the light reflected from the imaging target are measured as a function of axial depth
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A computer-implemented method of processing each of a plurality of sets of OCT images in a sequence of OCT images of a portion of a retina to generate a respective indication of a tissue velocity at a position along an axial direction in the OCT images, by performing, for each set: calculating a respective comparison value based on an image quality or similarity of a first OCT image and a second OCT image in the set; using phase information in the first OCT and second OCT image in a calculation of the respective indication of tissue velocity at the position if the comparison value is equal to or greater than a threshold; and omitting the phase information from the calculation if the comparison value is smaller than the threshold.