RESEARCH PLAN: VISION BIOSTATISTICS MODULE OVERVIEW The Vision Biostatistics Module will provide statistical expertise for the vision research community at UCSD. To date, each investigator has managed its data and analysis resources independently, often duplicating resources and expertise. Moreover, each investigator has not had adequate capability to acquire sufficient statistical resources to bring their research to the next level of productivity and sophistication. The Vision Biostatistics Module will provide statistical analyses and consultation to investigators of both clinical and experimental studies towards a common goal of improving the research productivity and expanding collaboration in the vision research community at UCSD. The staff senior biostatistician, the first to reside at the Department of Ophthalmology, will enhance the research capabilities of each investigator by providing dedicated statistical consultation and analyses for vision researchers. Currently, each investigator either has a collaborative relationship with an outside statistician or completes the statistical analysis himself or through students and/or postdoctoral fellows. Most often the consultant biostatistician is involved in a very limited part-time capacity, limiting his/her ability to develop an in-depth understanding of vision-specific analysis issues and to keep abreast of recent advancements in the field. There are several common analysis themes and statistical issues that can be addressed effectively and efficiently by having a dedicated biostatistician familiar with eye research to analyze vision related data. For example, there are statistical issues and novel methods that can be used to address the analysis of longitudinal imaging data of subjects/animals with two (non-independent) eyes contributing to the analysis. In addition, most of the grants (both human and animal) using this module utilize imaging instruments such as spectral domain optical coherence tomography to provide measures of retinal structure. These instruments provide a large amount of correlated measurements and summary parameters that must be analyzed efficiently with appropriate statistical models. The biostatistician in this module will become familiar with these summary measures and analytic challenges they represent which in turn will serve as a resource to other modules using these technologies. The familiarity of a dedicated statistician with vision research data should greatly enhance productivity over time as it will remove the time-consuming steps of explaining the data structure to outside statisticians. Finally, utilizing genetic, imaging and functional data to predict the development or progression of various eye diseases involves sophisticated prediction modeling techniques that are similar across grants.
Showing the most recent 10 out of 266 publications
