The goal is to develop precise, accurate, and efficient threshold estimation of the full visual field. We have used the Open Perimetry Initiative to develop custom perimetry tests which are available from github. The specific aims of this proposal are (1) to characterize the location of far peripheral visual field defects of optic neuropathies to static stimuli, (2) to develop a Bayesian strategy to test the full visual field in less than 10 minutes per eye, (3) to validate the full visual field testing with the new Bayesian strategy perimetry test, and (4) to determine whether the correlation between retinal nerve fiber layer structure from Optical Coherence Tomography and perimetric results with the full visual field is greater than using only the central visual field.
The long-term goal is to reduce the time needed to detect glaucoma progression. With an individualized model to detect progression that uses structural and functional data jointly, the specific aims are (1) to determine whether individualized approaches to identify glaucoma progression leads to earlier detection of progression compared to methods based on population statistics, (2) to determine which combination of structural and functional parameters identifies glaucoma progression at the earliest point in time, and (3) to determine the shortest period of time needed for our individualized approach to detect glaucoma progression.
The long-term objectives were to identify optical signals that control accommodation and emmetropization of the eye, and to identify the mechanisms that mediate the optical signals. We found that accommodation does not respond differently to different wavefront aberration characteristics. However, there is more to accommodation of the eye than simply minimizing retinal blur. Optical vergence with and without feedback was found to be an important cue for accommodation.
The long-term function of this research program was to relate visual deficits to the underlying cellular pathophysiology of disease processes, with the current focus exclusively on glaucoma. The research applies a quantitative cortical neural pooling model to analysis of perimetric damage produced by glaucoma, with the goals of reducing perimetric variability and improving relations between clinical measures of glaucomatous damage. The R package visualFields was originally developed as part of this project.
As part of the ongoing project, routine statistical analyses were developed, along with exploratory analyses of image quality metrics of the accommodating eye for
emmetropes and myopes.
This applications-oriented project aimed to provide a simple, flexible procedure for estimating psychometric functions and deriving thresholds from them, with minimum assumptions. The approach was based on locally weighted polynomial regression. The deliverable of the project was a package called modelfree.
The aims of the project were to determine the limits on how absolute and relative information about surface color is coded, how that information is affected by memory, and how current theories of surface-color coding need to be modified to account for actual human performance. We developed a MatLab package to compute the offset Kozachenko-Leonenko estimators of Shannon's differential entropy and of continuous mutual information.
The aims were to quantify how much information (in Shannon's sense) is carried by achromatic and chromatic pathways in viewing natural scenes under different illuminations and the effects of the different spatial resolutions of these two pathways on information-carrying capacity.
We validated the structure-function model in an independent cohort (the Ocular Hypertension Treatment Study) and estimate the sensitivity and specificity of the structure-function model for predicting which patients will progress.
We continued the development of the structure-function model and compared it against conventional models of glaucoma progression to see which model detected progression earlier.
We developed a model to improve the detection and monitoring of glaucoma progression. This model combined structural and functional data and was individualized to each patient. The model summarized the dynamics of structure-function change as the diseases progresses. It estimates the state of the disease (position and velocity) vectors.