MEDx is supporting five of the 25 Data+ 2018 teams. Data+ is a 10-week summer research experience for Duke undergrads interested in exploring new data-driven approaches to interdisciplinary challenges. Many of the data scientists are engineers helping to solve problems from the School of Medicine and Duke Health.
Teams are comprised of 2-3 undergraduate students and 1-2 graduate student mentors. Through this experience, students learn how to marshal, analyze and visualize data while gaining broad exposure to the modern world of data science. “This program harnesses the incredible talent of our undergrads and helps to rapidly solve real-world problems,” MEDx director, Geoff Ginsburg said.
Learn more about the five projects:
Improving the Machine Learning Pipeline at Duke
The team will compare and contrast conventional (Oracle Exadata) and distributed (Apache SPARK) systems in the analysis of EMR data and create recommendations for implementation. They will then use these systems to execute natural language processing on clinical narratives and radiology notes with existing, ongoing analyses of Duke data. Learn more
Rare Metabolic Diseases
The team will explore patterns of health care treatment and utilization for several rare metabolic disorders treated at Duke University Health System (DUHS). Students will interact with faculty experts from multiple disciplines to demonstrate how data-driven clinical profiles can inform our understanding of patients’ health care experience and support clinical care and research. Learn more
Deep Learning for Single Cell Analysis
The team will develop methods to identify cell subsets and their developmental, maturation and activation lineage relationships using deep learning approaches. Students will learn to process single cell RNA sequencing data and use the Python programming language and TensorFlow to characterize lung stem cells involved in wound healing. This work will help Duke researchers establish a deep learning pipeline for single cell analysis with applications in immunology, cell biology and cancer. Learn more
Big Data for Reproductive Health
The team will develop an interactive, web-based platform that curates raw data on contraceptive discontinuation from the Demographic and Health Surveys into a tool to help researchers and family planning advocates develop fresh insights around contraceptive discontinuation. Students will develop and refine the prototype, debut it with experts in online data visualization platforms at RTI and prepare a dissemination plan for the tool. Learn more
Complex Decisions, Real Numbers: Medical Decision-Making
The team will explore a plethora of physician-patient conversations and unravel the decision-making process. Students will be introduced not only to data science but also to behavioral research and aspects of communication in healthcare. This work will inform physicians on how to reduce overutilization of unnecessary interventions and ensure the well-being of patients. Learn more