October 7, 2020

Alex Quistberg, PhD, MPH, assistant research professor in the Dornsife School of Public Health Department of Environmental and Occupational Health and the Urban Health Collaborative (UHC) was recently awarded a 5-year, $690,000 K01 International Research Scientist Development Award from the Fogarty International Center at the National Institutes of Health.

The Built Environment, Pedestrian Injuries and Deep Learning (BEPIDL) study will focus on developing neighborhood built environment indicators and examining their association with pedestrian safety in Bogota, Colombia. Quistberg will lead a research team that will create the neighborhood indicators from built environment features labeled automatically in street images by neural networks.

A major challenge to improving road safety and other health outcomes in low and middle-income (LMIC) cities is data on built environment features that have important influences on health. Automating the collection of these data has the potential to drastically reduce the barriers to citywide built environment and traffic safety and other health research in LMIC cities, thus substantially increasing research capacity and generalizability.

His study team will also examine the equitable distribution of road safety features or dangers by neighborhood measures of the social environment.

This study will draw on the expertise of several investigators of the SALURBAL (Urban Health in Latin America) project, who will mentor Quistberg, including Dean Ana Diez Roux, MD, PhD, MPH, and Gina Lovasi, PhD, MPH, both from the Dornsife School of Public Health and the UHC, and Dr. Olga Lucia Sarmiento, MD, PhD, MPH, from Universidad de los Andes in Bogota, Colombia.

As part of the project, he will spend in-person time in Bogota collaborating with Sarmiento, as well as computer vision expert Pablo Arbelaez, PhD. He will also be mentored by Quynh Nguyen, PhD, MSPH, at University of Maryland in creating neighborhood indicators from images labeled by neural networks and will receive additional mentoring from Tolga Tasdizen, PhD, MS, at the University of Utah in developing neural networks for image analysis.