Visual Field Test Apparatus Using Statistical Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current visual field tests, such as the Moorfields Motion Displacement Test, face challenges with unreliable responses and prolonged test durations, leading to increased costs and subject fatigue, which affects the accuracy and reliability of sensitivity measurements.
Innovation Solution
A method and apparatus that utilize a statistical model to estimate sensitivity levels across a visual field by presenting a sequence of visual stimuli and incorporating binary responses, leveraging correlations between locations to reduce the number of queries and incorporate Bayesian approaches for efficient threshold determination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a sequence of multiple stimuli is presented at each location to account for unreliable responses, then the reliability of sensitivity measurements is improved, but the test duration increases
Solution Approach 1:
The patent applies partial action by presenting stimuli at only some locations at each time step rather than all locations. The system selectively presents stimuli based on current threshold estimates, focusing computational and testing resources on locations most likely to provide informative data, thereby reducing overall test duration while maintaining reliability through targeted repeated presentations at critical locations
Solution Approach 2:
The patent implements dynamics by adapting the stimulus presentation sequence based on real-time responses and updated threshold estimates. The system dynamically adjusts which locations receive stimuli next, modifying the test pathway based on accumulated information. This dynamic adaptation allows the test to converge faster by focusing on uncertain locations while skipping locations with already well-established thresholds
2Productivity
If the test duration is reduced to minimize subject fatigue and increase throughput, then productivity is improved, but the reliability of sensitivity measurements deteriorates
Solution Approach 1:
The patent applies feedback by using subject responses to continuously update threshold estimates, which then inform subsequent stimulus presentation decisions. The system monitors responses in real-time and adjusts the testing strategy based on this feedback, allowing it to identify reliable sensitivity measurements faster and reduce the number of locations requiring repeated testing, thereby maintaining reliability while improving throughput
Solution Approach 2:
The patent applies preliminary action by using initial responses to establish preliminary threshold estimates that guide subsequent testing. These preliminary estimates allow the system to prioritize locations needing further testing versus those with sufficient data, enabling faster convergence to reliable measurements and reducing overall test duration while maintaining measurement reliability
3Measurement precision
If traditional sequential testing methods are used to determine thresholds, then measurement precision is maintained, but the test duration increases due to lack of utilization of correlations between locations
Solution Approach 1:
The patent applies merging by combining information from responses at different locations to improve threshold estimation efficiency. The system uses a probabilistic framework that integrates responses across multiple locations, leveraging spatial correlations in visual field sensitivity. This allows the system to extract more information from each response, reducing the total number of stimuli needed while maintaining measurement precision through combined information from correlated locations
Data Source
AI summary
A method and apparatus are provided for measuring a sensitivity level across a visual field of a subject defined by a set of locations. The method comprises: presenting a sequence of visual stimuli on a display, wherein each stimulus in the sequence has a respective intensity level and is positioned on the display to correspond to a respective location from the set of locations; obtaining from the subject, for each stimulus, a respective binary response indicating whether or not the stimulus was seen by the subject; and after receiving the response from the subject for a given stimulus in the sequence, using a statistical model to estimate, for each location in the set of locations, the sensitivity level at that location; wherein the statistical model incorporates information about correlations in responses between different locations from the set of locations and takes as input, the respective binary response, intensity level and location for multiple stimuli presented in said sequence up to and including the given stimulus.