2018 ISIB Projects

  • Assessing Imaging Data from the Michael J. Fox Foundation’s Parkinson’s Progression Markers Initiative for use in Future Clinical Trials
    Parkinson’s disease is a progressive neurodegenerative disorder that affects movement.  The disease is characterized by a gradual worsening of symptoms such as shaking, stiffness, and difficulty with walking, balance, and coordination.  The gradual worsening of symptoms makes clinical research in Parkinson’s disease extremely difficult as studies often need to include huge numbers of patients and long durations of follow-up, making them extremely costly and ultimately not feasible.  The Michael J. Fox Foundation’s Parkinson’s Progression Markers Initiative (PPMI) is a prospective cohort study of de novo Parkinson’s patients and healthy controls that seeks to identify biomarkers that predict Parkinson’s progression.  A predictive biomarker would allow for shorter and smaller studies, speeding up the treatment discovery process immensely.  One potential biomarker is DaTSCAN, a type of imaging that is used in the diagnosis of Parkinson’s disease.  This project seeks to model the PPMI DaTSCAN data and to make projections about the sample size and duration that a future study would need in order to use DaTSCAN as a biomarker for potential Parkinson’s disease-modifying treatments.
  • An Analysis of the Efficacy of State Policies Designed to Address Bullying Victimization
    Youth violence in the United States is common and is associated with numerous adverse health, academic and psychosocial outcomes. Programs aimed at preventing youth violence have shown only modest success. Consequently, researchers have begun exploring additional primary prevention strategies that can augment existing programs, including anti-bullying policies. Though all 50 states have enacted some type of anti-bullying legislation, there is a striking dearth of research on whether, why, and for whom these laws are effective in preventing bullying and other forms of youth violence. In order to begin to address this deficiency, we will use longitudinal state bullying data supplied by the Youth Risk Behavioral Surveillance System (YRBSS) and preliminary legal data obtained from the first longitudinal content analysis of anti-bullying laws (and amendments) from their inception in 1999 through 2017 to investigate the role of policy in reducing bullying victimization.
  • Radiomics for Disease Characterization and Outcome Prediction in Cancer Patients
    Radiomics is an emerging field of medical study in which large numbers of quantitative features are extracted from medical images of patients to aid in the characterization of disease and prediction of clinical outcomes.  This project will focus on the application of radiomics to cancer patients.  Project participants will get hands-on experience with the different steps involved in radiomics analysis, including medical image visualization, tumor segmentation, feature extraction and quantification, and statistical analysis.  Patient images will be visualized with the 3D Slicer open-source software.  Segmentation software tools will be presented and applied to identify regions in the images that contain tumors.  From the segmented regions, quantitative features will be calculated as measures of the volume, shape, activity, and texture of the tumors.  Statistical analysis will then be applied to the resulting radiomic features in combination with other clinical information on patients.  In particular, regression and machine learning techniques will be employed to identify important features, develop multivariable statistical models to predict clinical outcomes, and evaluate the performance of the models.
  • Understanding the Opioid Crisis in Iowa through Data
    The opioid crisis has been a focus of much research and media attention over the last few years in response to worrying trends. In 2011, opioid overdoses became the leading cause of injury related mortality in the United States, and sales of opioid pain relievers (OPRs) have grown dramatically since the early 2000’s. While both legally prescribed and illegally obtained opioid drugs are associated with overdose deaths, recent research indicates that legally prescribed opioid pain relievers are an important cause of dependence, which can in turn lead to abuse and even overdose. In this project, we will examine OPR prescription patterns in the state of Iowa, and evaluate associations with medical diagnoses of abuse and dependence. This work will be conducted through the analysis of private medical insurance claims.
  • The Role of Lithium in Suicide Prevention
    In the U.S., 4.6% of people report having attempted suicide at least once, and 1.2% of all deaths are attributed to suicide.  Both genetic and environmental factors are believed to play a significant role in the risk for suicide.  Taking lithium has been shown to lower suicidal tendencies, although the exact mechanism by which it accomplishes this is unknown, as is the reason lithium prevents suicides in some individuals but not in others.  Researchers at the University of Iowa have recently treated cell lines with lithium and used RNA-Seq to measure the genome-wide effects of acute and chronic lithium treatment on cells.  The hope is to cross-reference these results with genes and pathways identified in other genetic studies of suicide in order to better understand how lithium works to decrease suicidal behavior.
  • Understanding the “Who”, “Where”, and “When” of Clubfoot in Iowa
    The Iowa Registry for Congenital and Inherited Disorders (IRCID) conducts statewide surveillance for major structural birth defects diagnosed among pregnancies of Iowa residents. Idiopathic talipes equinovarus, more typically referred to as clubfoot, is a common musculoskeletal birth defect. Despite nearly one-half century of study, major risk factors for this defect (other than perhaps cigarette smoking during pregnancy) remain elusive. Reports of prevalence estimates for clubfoot tend to be higher among Hawaiians and Maoris, intermediate for Caucasians, and lowest for African Americans. These estimates vary by geographic region, owing perhaps to differences in racial/ethnic distributions in these regions, but also perhaps due to other unidentified risk factors related to environmental exposures, socioeconomic factors, neighborhood effects, etc. Using IRCID data and following the common reporting paradigm for public health surveillance that focuses on the information triad – person, place, and time – this project will investigate trends in prevalence of clubfoot in Iowa over a 20-year birth period using spatial statistical modeling to identify geographic hotspots of clubfoot occurrence. Multilevel Poisson regression analyses and Bayesian methods will be used for the proposed analyses.
  • Predicting Clostridium difficile Infections in Healthcare Facilities with a Focus on Testing Potential Interventions
    According to the CDC, Clostridium difficile (C.difficile) caused almost half a million infections in the U.S. alone, directly leading to 15,000 deaths annually.  This bacterium is listed as an “Urgent Threat” not only due to its prevalence but also its antibiotic resistance.  People can become infected if they touch contaminated surfaces and then touch their mouth, nose, eyes, etc.  There are a variety of risk factors for C.difficile infections, such as age, length of hospital stay, and antimicrobial use.  It is important to be able to predict when C.difficile rates may pose a high risk to patients coming into the hospital for unrelated morbidities, as well as which factors contribute most significantly to this higher risk.  With an accurate prediction model, we may then test various interventions via a “virtual laboratory” to determine an optimally efficacious strategy.
  • Risk Factors for Death by Suicide and Homicide in Young Americans
    According to the Centers for Disease Control, suicide and homicide are ranked second and third among leading causes of death in the 15-24 year age group in the U.S.  This project will apply methods from statistics and machine learning to state- and county-specific data to determine what demographic, social, and governmental characteristics are associated with increased risk of death by suicide or homicide in young Americans.