Eye Inhomogeneity Characterization Using OCT and Camera Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing ophthalmological diagnostic systems for characterizing optical inhomogeneities in the eye, such as cataracts, floaters, and corneal opacification, lack robustness and accuracy, particularly in grading cataracts, and do not effectively utilize high-resolution data from optical coherence tomography (OCT) imaging.
Innovation Solution
A computer-implemented method and device that combines optical coherence tomography data with complementary imaging techniques like direct illumination, retroillumination, and Scheimpflug imaging, using machine learning models, specifically convolutional neural networks, to characterize optical inhomogeneities by identifying area and volume segments, and adapt treatment patterns for laser treatment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If OCT-based measurements are used to classify cataracts, then acquisition time is reduced and resolution is improved, but it is unclear whether OCT images alone contain sufficient information for accurate grading
Solution Approach 1:
The patent combines OCT images with slit lamp images and retroillumination images to create a comprehensive diagnostic system. By merging multiple imaging modalities, the system leverages the high resolution of OCT while supplementing it with the diagnostic information from slit lamp and retroillumination imaging, thereby achieving both high measurement precision and reliable cataract grading.
Solution Approach 2:
The patent introduces a machine learning model as an intermediary that processes and integrates data from multiple imaging sources. The model acts as a mediator that combines the strengths of different imaging techniques and produces accurate cataract classifications, resolving the uncertainty about whether OCT alone is sufficient for grading.
2Reliability
If multiple imaging techniques are combined, then detection confidence and accuracy are enhanced, but device complexity increases
Solution Approach 1:
The patent implements a multi-functional imaging system that can perform OCT imaging, slit lamp imaging, and retroillumination imaging using a single integrated device. This universal approach allows the system to gather comprehensive diagnostic information while managing device complexity through consolidation rather than separate independent systems.
Solution Approach 2:
The machine learning model automatically processes and integrates data from multiple imaging techniques without requiring manual intervention. The system self-services by autonomously combining the imaging data and producing diagnostic results, thereby managing complexity through automation rather than manual multi-system coordination.
3Measurement precision
If traditional LOCS III classification is used, then cataract grading is accurate, but it requires special training and additional labor-intensive steps
Solution Approach 1:
The patent replaces the manual mechanical process of LOCS III classification with an automated machine learning system. The machine learning model processes imaging data and performs cataract grading automatically, substituting the labor-intensive manual classification process while maintaining or improving grading accuracy and significantly increasing diagnostic efficiency.
Solution Approach 2:
The machine learning model performs cataract classification autonomously without requiring ophthalmologists to undergo special training or perform manual grading steps. The system self-services by automatically analyzing imaging data and producing classifications, thereby eliminating the need for specialized training and reducing labor-intensive procedures while maintaining accurate grading.
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
Enhances detection confidence and accuracy in classifying cataracts and other optical inhomogeneities, providing a more comprehensive understanding of the eye's inhomogeneities for precise surgical planning and treatment.
Implementation Method 1
Optical coherence tomography provides a method for creating an optical cross-section of the eye which is non-invasive and high-resolution. The type of light used in OCT allows it to penetrate deep into tissue and examine tissue and fluid layers even with a high reflectivity.
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
A computer-implemented method and device for characterizing an optical inhomogeneity in a human eye is disclosed, the method comprising: receiving (S1) optical coherence tomography data of the eye; receiving (S2) image data of an image of the eye recorded by a camera, the image recorded using one or more of the following imaging techniques: direct illumination of the eye, retro illumination of the eye, or Scheimpflug imaging; and characterizing (S3) the optical inhomogeneity as one or more of the following optical inhomogeneity types: a cataract, a floater, or an opacification of the cornea using the optical coherence tomography data and the image data.