Retina OCT Image Analysis for Diffuse Intraretinal Fluid Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Optical Coherence Tomography (OCT) images struggle to reliably identify and quantify diffuse intraretinal fluid (DIRF), which is crucial for assessing macular oedema and guiding treatment, due to low accuracy in segmentation techniques.
Innovation Solution
A method using a processor to segment boundaries between retina layers, determine layer locations, and analyze regions of pathology, including DIRF, to derive a comprehensive assessment of the retina, enabling accurate volume measurement and progress tracking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional segmentation techniques are used to analyze OCT images, then the analysis process is simple and fast, but the accuracy in identifying and quantifying diffuse intraretinal fluid is low
Solution Approach 1:
The patent applies segmentation by dividing the retinal image analysis into distinct layers and regions. The system segments the retina into multiple layers (nerve fiber layer, ganglion cell layer, inner plexiform layer, inner nuclear layer, outer plexiform layer, outer nuclear layer, photoreceptor layer, retinal pigment epithelium) and identifies specific pathology regions (focal intraretinal fluid, diffuse intraretinal fluid, subretinal fluid). This multi-level segmentation enables precise identification and quantification of DIRF that conventional single-stage segmentation cannot achieve.
Solution Approach 2:
The patent transitions from 2D cross-sectional OCT images to 3D volumetric analysis by acquiring multiple B-scans across the macula and reconstructing them into a three-dimensional dataset. This dimensional change allows for accurate volume measurement of diffuse intraretinal fluid and provides comprehensive spatial context that improves measurement precision beyond what single 2D slices can provide.
2Measurement precision
If detailed layer segmentation is performed for each pixel, then measurement precision of fluid volume is improved, but processing time and computational complexity increase
Solution Approach 1:
The patent performs preliminary action by first segmenting the retinal layers and identifying pathology regions before conducting volume measurements. The system pre-processes the 3D OCT data to create segmented masks of different retinal layers and fluid regions, which are then used for efficient volume calculation. This preliminary segmentation step organizes the complex data structure in advance, making subsequent measurements faster and more accurate.
Solution Approach 2:
The patent replaces traditional mechanical/image processing methods with automated computer-based analysis. The system uses algorithms to automatically segment layers, identify pathology regions, and calculate volumes from 3D OCT datasets, substituting manual measurement methods with computational approaches that provide both high precision and efficient processing of large volumetric datasets.
Data Source
AI summary
Provided is a method and system for analysing images of a retina captured by an Optical Coherence Tomography (OCT) scanner. In some examples, an image of a retina of a patient from an OCT scanner is received; boundaries between layers are segmented for each of the pixels; the layers are determined using the segmented boundaries; regions of pathology are segmented for each of the pixels; a location of the regions of pathology are determined with respect to the determined layers using the segmented regions; the regions of pathology are determined using the segmented regions and the determined location of the regions; a property of the regions of pathology are determined using the segmented regions; results of determinations of the regions of pathology and the property of the regions to derive an assessment are analysed; and the assessment are output.