Wilkins, Mark; (2002) Visual field changes following trabeculectomy. Doctoral thesis (M.D), UCL (University College London).

Text Visual_field_changes_following.pdf Download (14MB)

Abstract

The use of automated visual fields to detect and monitor glaucoma is hampered by having no gold standard against which to compare them. In the case of monitoring disease progression visual fields display large amounts of fluctuation that can mask true change. The analysis of fields using pointwise linear regression (PLR) has been developed to more accurately detect change. However the criteria for change using PLR are themselves poorly understood. This thesis examines the collection of field data from a surgical trial of trabeculectomy and then explores the detection of change in the eyes in the study using conventional and PLR grading techniques. Analysis of field data from an initial group of patients in the trial reveals the large amount of change detected using existing criteria. Much of the change detected is due to noise or fluctuations in a patient's response that do not represent real change. The use of modified criteria has variable effects on the detection of change. From this group of modified criteria, 6 can be selected on an empirical basis. All maximise the detection of progression while minimising improvement. Given the data available it is not possible to link any changes in visual field to changes in media opacity, especially cataract. When the selected criteria are tested against a) extended follow up data and b) a second group of patients from the same trial one criterion offers the ability to detect progression in both groups of patients while minimising the detection of improvement. This criterion requires a particular spatial arrangement of points in the field. Analysing groups of patients' fields using PLR without regard to treatment offers a way of developing change criteria prior to analysis within treatment arms.

Download activity - last month

Export as XMLJSONCSV

Download activity - last 12 months

Export as XMLJSONCSV

Downloads by country - last 12 months

Export as CSVJSONXML

Archive Staff Only

View Item