OCT Scan Analysis for Early Geographic Atrophy Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods fail to accurately identify geographic atrophy (GA) in age-related macular degeneration until symptoms appear, limiting the ability to target subjects for clinical trials and prevent advanced disease.
Innovation Solution
A system using machine learning with a trained GA detection algorithm analyzes OCT volume scan images to generate a probabilistic likelihood of GA presence or onset, utilizing a computing platform with a processor and memory to determine GA presence or likelihood through a machine learning algorithm trained on relevant datasets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional diagnostic methods are used, then GA can be identified, but only after symptoms appear and disease has progressed
Solution Approach 1:
The machine learning algorithm performs preliminary analysis of OCT images to detect early signs of geographic atrophy before clinical symptoms manifest. By analyzing structural changes in the retinal pigment epithelium and underlying layers, the system identifies pre-symptomatic GA, enabling early intervention before irreversible vision loss occurs.
Solution Approach 2:
The patent replaces conventional manual diagnostic methods with an automated machine learning-based image analysis system. The algorithm processes OCT volumetric data to generate probabilistic assessments of GA presence, substituting human visual inspection with computational analysis that provides more consistent and earlier detection.
2Loss of information
If conventional diagnostic approaches are used, then GA detection is possible, but subject targeting for clinical trials remains imprecise
Solution Approach 1:
The system replaces manual diagnostic assessment with automated machine learning analysis of OCT images. The algorithm processes volumetric retinal data to generate objective probabilistic assessments of GA presence and likelihood, providing quantifiable information that improves subject identification accuracy for clinical trials.
Solution Approach 2:
The machine learning algorithm acts as an intermediary between raw OCT imaging data and clinical diagnosis. It processes complex volumetric image data to extract subtle structural features indicative of GA, translating imaging data into actionable diagnostic information that bridges the gap between imaging and clinical decision-making.
Data Source
AI summary
Provided herein are systems utilizing machine learning in detecting geographic atrophy. The system can include at least one processor, a memory, and a computing platform including the at least one processor and the memory. The computing platform can be configured to receive volumetric scan images, determine a position of each of the volumetric scan images, generate, using a trained detection algorithm, a probabilistic likelihood that each scan is informative of geographic atrophy or neovascular age-related macular degeneration, include high probability scans in a dataset, determine, using the included scans in the dataset, whether geographic atrophy or neovascular age-related macular degeneration is present or likely to occur, and output, by the detection algorithm, information indicating whether or not geographic atrophy or neovascular age-related macular degeneration is present or likely to occur.