Spatial Modeling in Glaucoma

Ohio State University, Columbus OH

Investigators

Linked publications & trials

Abstract

DESCRIPTION (provided by applicant): This career training proposal is to train Michael D. Twa, OD, MS as an independent clinician-scientist. A five year training program is proposed, consisting of formal coursework in vision science, specific training in computer science and image processing, and mentoring in the application of these skills to clinical outcomes research in glaucoma. In September 2003, NIH announced a new "Roadmap" to accelerate advances in biomedical research for the 21st century. Three areas listed in this Roadmap are relevant to this research proposal: (1) Interdisciplinary research training. (2) Clinical research informatics. (3) Development of enabling technologies for improved assessment of clinical outcomes. The Roadmap emphasizes coordinated strategies to develop both technological and human resources to take full advantage of multidisciplinary and translational research opportunities. This proposal addresses the stated training objectives at an individual level. Glaucoma is a leading cause of blindness. Visual field assessment and optic nerve head imaging (confocal scanning laser tomography) are commonly used to diagnose the disease and monitor its progression, yet there is considerable controversy about how to interpret and make best use of this information. Currently, raw data from these observations are reduced to statistical indices that are meant to summarize clinically meaningful features and provide a basis for classifying test results as normal or not. Unfortunately, these indices may sacrifice other relevant features in the data for interpretability. We will use mathematical modeling methods (polynomial modeling, spline fitting and wavelet analysis) to quantify patterns in visual field data and topographic images of the optic nerve head. We will use features derived from these modeling methods to apply novel pattern recognition techniques from computer and information sciences-decision trees and non-linear regression analysis-and then compare these techniques to current methods to identify glaucoma. By improving current methods of analysis we can provide a more quantitative basis for clinical decisions, and offer greater consistency and objectivity on data interpretation. The long-term objective of this proposal is to translate advances in computer and information sciences to the analysis of clinical outcomes research in glaucoma and other eye diseases.