Retinal Blink Rejection Using Filtered SLO Image Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for detecting eye blinks during retinal imaging are ineffective due to low information content and similar intensity characteristics between blink and non-blink conditions, particularly in peripheral scans, leading to unclear images and compromised scan quality.
Innovation Solution
A method involving spatial directional filtering and object filtering of IR SLO images, followed by cross-correlation to distinguish between blink and non-blink conditions, allowing for the removal of unclear images and improved retinal tracking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional blink detection methods are used, then the detection process is simple, but the detection accuracy is low due to similar intensity characteristics between blink and non-blink conditions
Solution Approach 1:
The image processing is divided into multiple stages: spatial directional filtering to enhance edges, object filtering to identify and remove non-retinal objects, and cross-correlation to compare processed images. This segmentation allows each stage to focus on specific features, improving blink detection accuracy while managing complexity through modular processing
Solution Approach 2:
The patent introduces intermediate processed images as mediators between the raw images and the final blink detection result. The spatially filtered images and object-filtered images serve as intermediate representations that highlight relevant features and remove distractions, enabling more accurate cross-correlation-based blink detection
2Reliability
If all captured images are included in the final frame stack, then the image acquisition is complete, but the final image quality is reduced due to inclusion of unclear images from blink conditions
Solution Approach 1:
The system performs preliminary processing of images during acquisition, applying spatial filtering and object filtering to identify unclear images before they are included in the final frame stack. This preliminary action allows the system to pre-screen images for quality, ensuring that only clear images are used in the final reconstruction, thereby maintaining high reliability without sacrificing productivity
Solution Approach 2:
The cross-correlation process provides feedback about image quality by comparing each captured image with reference images. Images with low correlation scores are identified as unclear and excluded from the final frame stack, creating a feedback mechanism that automatically quality-controls the acquired images and maintains high final image reliability
3Measurement precision
If retina tracking is performed without blink detection, then the tracking system is simpler, but the scan position accuracy is compromised due to loss of correct position during blinks
Solution Approach 1:
The tracking system is segmented into independent components: image acquisition, spatial filtering, object filtering, cross-correlation for position determination, and blink detection. This segmentation allows the system to accurately determine scan position only when clear images are available, maintaining high position accuracy while managing complexity through modular, condition-based processing
Solution Approach 2:
The system performs preliminary blink detection and exclusion of unclear images before retina tracking is performed. By preemptively identifying and removing images captured during blinks, the system prevents position corruption from occurring, maintaining accurate tracking without requiring complex real-time correction mechanisms during the tracking process itself
Data Source
Figure 1
Figure 2
Figure 3a~3c
AI summary
A method, system, and computer-readable medium, for detecting whether an eye blink or non-blink is captured in the image. The method includes filtering, from the image, one or more objects that are predicted to be unsuitable for determining whether an eye blink or no-blink is captured in the image, to provide a filtered image. The method also includes correlating the filtered image with a reference image, and determining, based on the correlating, whether the eye blink or non-blink is captured in the image. The eye blink is a full eye blink or a partial eye blink, and the images may be sequentially captured IR SLO images, in one example embodiment herein. Images determined to include an eye blink can be omitted from inclusion in a final (e.g., OCT) image.