Optometry and Vision Sciences - Theses
Permanent URI for this collection
Search Results
1 - 1 of 1
-
ItemMeasuring central visual field loss using visual stimuli with natural scene statistics( 2021)Routine visual field assessments are performed using static automated perimetry that measures contrast detection performance for white luminance increment targets presented on a uniform luminance background. While offering a standardised measure of the visual field, such clinical visual field tests are not representative of natural viewing conditions. Previous studies show that some visual field defects are detectable via visual search behaviour analysis; for example, when watching a video or performing daily living tasks. Such measurements of performance using visual stimuli representing everyday tasks can provide information regarding how people with visual field loss perform day to day activities, but due to widely varying visual content, it is difficult to predict a generalised explanation of performance with these tasks to create a widely applicable screening tool. This thesis aimed to develop a method for detecting central visual field loss by measuring the number of fixations to find targets on a background with spatial frequency content similar to natural scenes (referred to as 1/f noise). An advantage of this proposed approach is the existence of an established theoretical framework [Najemnik, J., & Geisler, W.S. (2005), Nature, 434(7031), 387-391] that links the detectability of targets within an image to the number of fixations required by a Bayesian ideal observer to find the targets. The stimulus and behavioural methods used in this thesis were chosen to be consistent with those used to validate the Bayesian ideal observer model. Four experiments were explored in this thesis that stepped through creating the screening protocol for detecting central visual loss. In the first step, Experiment One computationally assessed the suitability of using similar stimuli and behavioural methods as Najemnik and Geisler (2005) to detect field loss. Experiment Two evaluated the need for an age-matched normative limit for setting the contrast of the target for the screening protocol using such a method. Experiment Three determined the requirement of age-matched normative data for the number of the fixations to find the targets on the 1/f noise background, the outcome measure for the screening protocol. The outcomes from Experiment One to Three were then applied to Experiment Four that developed a prototype test for detecting central visual field loss by measuring the number of fixations required to search for a target on a 1/f noise background. This experiment also tested if the developed prototype was able to detect central visual field loss in a group of glaucoma participants. The prototype test demonstrated 85% sensitivity for a fixed specificity of 95.2% in screening abnormal areas in central vision in the glaucoma group. The task requirements of the screening protocol were designed to be representative of natural visual environments and to be intuitive for participants to perform the test. While the prototype test was tested in participants with glaucoma, the developed methods in this experiment are designed to generalise to any form of central visual field loss.