AI Glaucoma Screening Using Time-Stamped OCT Scan Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for screening glaucoma are inefficient due to reliance on single intraocular pressure measurements, which are inaccurate for 'low pressure glaucoma' and ineffective for identifying at-risk individuals, compounded by a shortage of trained eye care professionals and incompatibility issues with OCT scan equipment.
Innovation Solution
A computer-implemented system that links time-stamped OCT scan data to a biometric identifier, using AI to analyze differences between scans over time to diagnose eye diseases like glaucoma, allowing for accurate and efficient screening across disparate locations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a single OCT scan is performed for screening, then the exam is simple and quick, but the accuracy is insufficient due to inability to detect rate of nerve fiber layer loss
Solution Approach 1:
The system performs periodic OCT scanning at multiple time points (baseline scan, follow-up scans at intervals) to capture temporal changes in the nerve fiber layer. This periodic measurement approach enables detection of the rate of loss over time, which is the key indicator for glaucoma diagnosis, while maintaining efficient automated screening processes.
Solution Approach 2:
The system performs a baseline OCT scan and establishes reference measurements before the actual screening decision is made. These preliminary baseline measurements are stored and used for comparison with subsequent scans, enabling accurate detection of progressive nerve fiber layer loss without requiring complex real-time analysis during each individual scan.
2Measurement precision
If serial OCT measurements are performed, then accuracy improves through detection of changes over time, but practical obstacles arise due to patient practice changes and equipment incompatibility
Solution Approach 1:
The system uses a standardized data exchange format and biometric identifier as an intermediary between different OCT equipment manufacturers and practices. This intermediary layer enables seamless transfer and comparison of OCT scan data across incompatible systems, allowing serial measurements to be performed and compared even when patients change practices or when different equipment manufacturers are involved.
Solution Approach 2:
The system implements a universal biometric identifier system that works across all participating practices and equipment types. This universal identifier enables the system to track patients across different practices and equipment, making the serial measurement system adaptable to various organizational structures and equipment configurations without requiring practice-specific implementations.
3Ease of operation
If intraocular pressure measurement is used for screening, then the screening process is simple, but the effectiveness is very low due to normal pressure in many glaucoma patients
Solution Approach 1:
The system replaces the mechanical tonometry method (intraocular pressure measurement) with optical coherence tomography imaging to measure nerve fiber layer thickness. This substitution eliminates the fundamental limitation of pressure-based screening while maintaining an automated, efficient screening process. The OCT-based measurement directly visualizes the structural damage characteristic of glaucoma, providing reliable detection independent of pressure levels.
Data Source
AI summary
Computer-implemented systems and methods link time-stamped OCT scan data of a patient's retina in a database to a biometric identifier for the patient. An appropriately trained artificial intelligence (AI) computer system determines whether the patient has an eye disease based on the differences between time-stamped OCT scan data for the patient from different time scans.