Biostatistics Summer Initiative

2009 Project List

Mentors for the projects will be:

Kathryn Chaloner, Jacob Oleson, Brian Smith, Gideon Zamba

Project: Ultrasound studies of early indicators of atherosclerosis – Thickened arteries in the chest and neck are known to be associated with heart disease in the elderly. We are currently using ultrasound technology and automated software to view and measure the thickness of artery walls in teenage, young adult, and middle-aged subjects. More work is needed in assessing the reliability and validity of such measures.

Project: Time-series modeling of driving studies – In driving simulators and in instrumented vehicles, we have been obtaining digitized information on the driving performance of healthy and impaired subjects. We are currently developing sophisticated mathematical models for the analysis of such data.

Project: Lung Airways tree assessment study — It is important to study the quality of the airways tree in normal human subjects, and set the study results as gold standard in order to identify smoking related diseases and cancerous cells that can damage human lung composition. To asses the quality of the lung, one uses the lung tree luminal areas and the relative wall thickness. However, there is a biological gender difference in lung composition as well as lung volume. This ongoing project will build a model for normal subjects’ lung composition and identify differences/similarities between males and females lung composition at various locations of the tree for the same lung volume.

Project: Cochlear Implant study — Cochlear implants have been designed to support speech recognition by persons with profound hearing loss. The Hybrid implant was developed at the University of Iowa and a clinical trial is underway with data collected over a three year period. We will assess the effectiveness of implants in children and adults on various measures.  Sound localization tasks (Speakers setup around the room and patients identify which speaker the sound came from), music perception and appraisal (How well do implant patients understand and appreciate various aspects of music?) and speech development (children with hearing aids and implants are assessed on speech development) are some of the areas that can be explored.  

Project: Disease Mapping — Disease mapping is commonly used to describe the variation in health outcomes over geographic regions. We will develop a model that can be used to obtain reliable estimates in rare diseases by incorporating information from more common diseases for which there exists a shared set of important risk factors. The spatio-temporal model will estimate relative risk, adjusted for age and gender, at the county level in Iowa for the period 1973--2006.

Project: Statistical Computing for Spatial Data Analysis – Spatial analysis allows for the mapping of environmental health effects, like particulate matter and residential radon that exhibit correlation over geographic regions. The large datasets involved in such analyses offer many interesting statistical and computational challenges. We are currently developing software for the analysis of spatial data. This ongoing project offers opportunities to learn about the graphical display and analysis of such data using the R statistical software program.

Project: The interaction of HIV and GB virus type C – Clinical and laboratory studies have indicated that the HIV virus interacts with a virus called GB virus type C. GB virus type C is transmitted in the same way as HIV, but is not associated with any known disease. Several epidemiological studies have shown that subjects HIV infection and also with GB virus type C infection live longer than subjects with HIV but without GB virus type C infection. Laboratory studies have indicated that there is an interaction between HIV and GH virus type C. This project will examine data on immunologic markers of subjects with HIV infection, some of whom are also infected with GB virus type C. The data analysis will use graphical exploration of multivariate data in the programs R and Ggobi.

Project: Genome-wide gene-mapping study of human disorders – The rapid development of genotyping technology makes it possible to search for genetic risk factors underlying human disorders over the whole genome by making use of high density single nucleotide polymorphisms (SNPs). The latest Affymetrix genome-wide human SNP array 6.0 can provide genotypes on up to 1 million SNPs. We have several genome-wide association studies and genome-wide linkage analyses going on that involve various kinds of human disorders. The data analysis will use genetics programs easyLinkage and PLINK.

Project: Modeling Temporal and Spatial effects of Glaucoma – Current methods for assessing visual fields and changes they undergo in time rely on the assumption that the various visual fields are independent. This assumption however is contrary to reality. Given that a location is affected by a disease, the chance for a neighborhood adjacent area to be affected is very high. This likelihood is even higher when it comes to glaucoma disease; because glaucoma seems to progressing from its starting location according to a pattern rather than scattering throughout the visual fields. This study will use a statistical model to estimate the rate at which normal subjects can loose their visual acuity in time, and use the normal model to assess visual loss progression in glaucoma patients on six progression patterns of the ocular globe: temporal, supero-temporal, infero temporal; nasal, supero-nasal, and infero-nasal. The study will further examine the spatial relationship within the various locations of the visual field.

Project: Stimulus Response Latency Estimation – Stimulus response latency is the delay between stimulus onset and the evoked modulation in neural activity. There is no cutting-edge technology in estimating this latency. A common way to estimate latencies involves binning spike arrival times to form a peri-stimulus histogram, smooth the histogram and choose the first time following stimulus onset in which the histogram exceeds the midpoint between its maximum and minimum value. This technique proves to be inaccurate most of the times. This project will explore the latency estimation by applying change-point methods to the empirical distribution of the spike arrival times, and compare the change-point method to the peri-stimulus histogram approach.  

Project: The Iowa Care for Yourself / Wisewoman Cardiovascular Study – The Iowa WISEWOMAN study is a program that provides cardiovascular screening and nutrition, physical activity, and smoking cessation interventions to low-income, under- and uninsured women aged 40 - 64 years.  The sessions focused on improving nutrition knowledge and behaviors, increasing physical activity, and building self-efficacy for health behavior change to reduce cardiovascular disease risk.  This is a longitudinal study where women are measured repeatedly over a period of 3 - 5 years.  We will study each of the risk factors to determine how health behavior patterns change over time for this population and what aspects of the program have made the greatest impact.

Return to Summer Initiative Program Information