Deep-Learning Ocular Landmark Measurement With Image Refinement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional methods for measuring landmark dimensions in the human eye, such as anterior chamber depth and lens thickness, are highly dependent on the surgeon's skill set, lack accuracy, speed, and repeatability, and rely on manual box thresholding and hard-coding, which are suboptimal.
Innovation Solution
Utilizing a deep-learning convolutional neural network (CNN) to autonomously detect landmark features in eye images, followed by real-time refinement using classical image processing techniques to enhance accuracy and reduce noise, enabling automated generation of biometric landmark dimensional measurements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual box thresholding and hard-coding methods are used to measure landmark dimensions, then the process can be performed with simple equipment, but the accuracy, speed, and repeatability of measurements deteriorate
Solution Approach 1:
The patent replaces manual mechanical measurement methods (box thresholding and hard-coding) with an automated deep learning system using convolutional neural networks. The CNN automatically detects landmark features and calculates dimensions, eliminating the need for manual intervention while significantly improving measurement accuracy, speed, and repeatability.
Solution Approach 2:
The deep learning system performs self-service by automatically detecting landmark points and computing dimensional measurements without requiring surgeon intervention. The system independently processes images, identifies features, and generates measurement results, reducing dependency on human skill while maintaining simple equipment requirements.
2Reliability
If deep-learning algorithms are used to autonomously detect landmark features, then measurement accuracy and repeatability improve, but the computational complexity and processing time increase
Solution Approach 1:
The patent replaces complex manual measurement procedures with a deep learning algorithm that automatically performs feature detection and measurement. The convolutional neural network learns to identify landmark features directly from images, providing consistent and repeatable results without the complexity of manual box thresholding and hard-coding procedures.
Solution Approach 2:
The system performs preliminary action by pre-training the convolutional neural network on labeled datasets before deployment. This preliminary training phase enables the algorithm to automatically recognize landmark features in new images, ensuring high repeatability and reliability without requiring complex real-time adjustments during actual measurements.
3Productivity
If manual landmark marking is performed by surgeons, then the measurement process can be simplified, but the speed and accuracy deteriorate due to dependency on surgeon skill
Solution Approach 1:
The system performs self-service by automatically detecting landmark features and computing measurements without requiring surgeon intervention. The deep learning algorithm independently processes images and generates results, eliminating the need for manual marking while improving both speed and accuracy regardless of surgeon skill level.
Solution Approach 2:
The patent replaces the manual surgical marking process with an automated computer-based deep learning system. This substitution eliminates the dependency on surgeon skill and manual dexterity, providing consistent high-speed measurements that are not limited by human factors.
4Measurement precision
If classical image processing techniques are applied after deep learning, then noise in dimensional estimates is reduced, but the overall processing time increases
Solution Approach 1:
The system applies feedback by using classical image processing techniques to refine the output of the deep learning algorithm. The post-processing step analyzes the initial measurements and adjusts them to reduce noise and improve accuracy, ensuring high-quality results while maintaining efficient processing through optimized algorithms.
Solution Approach 2:
The system applies partial action by selectively applying classical image processing techniques only to specific regions or measurements that benefit from refinement. This approach reduces noise in critical dimensional estimates without unnecessarily processing all image data, thereby minimizing additional processing time while maintaining high accuracy where it matters most.
Data Source
AI summary
A method for estimating biometric landmark dimensional measurements of a human eye includes, in a possible embodiment, receiving one or more images of the human eye via a host computer. In response to receiving the one or more images, the method includes generating a preliminary set of landmark point locations in the one or more images via the host computer using a deep-learning algorithm, and then refining the preliminary set of landmark point locations using a post-hoc processing routine of the host computer to thereby generate a final set of estimated landmark point locations. Additionally, the biometric landmark dimensional measurements are automatically generated via the host computer using the final set of estimated landmark point locations. A data set is then output that is inclusive of the set of estimated landmark point locations. A host computer that executes instructions from memory to perform the method.