March 28 @ 4:00 – 5:00 pm
Speakers: Javed Mostafa, PhD & Deborah Travers, PhD, RN, FAEN
Presented from UNC-CH
Broadcast Link: Seminar
Drs. Mostafa and Travers will describe approaches to automated syndrome classification using clinical text, as applied to a new system created to classify emergency department (ED) records that meet criteria for acute infectious disease surveillance (aka biosurveillance). The new system was created as part of an NLM-funded project, “Adapting NLP Tools for Biosurveillance.” The system is currently being evaluated for biosurveillance use by public health officials through the North Carolina Disease Event Tracking and Epidemiologic Collection Tool (NC DETECT). The presentation will include an introduction to NC DETECT, biosurveillance, and information retrieval techniques for data representation and classification, as well as a review of methods for evaluating automated classifiers. The new system, Emergency Medical Text Classifier (EMT-C) will be described, followed by a presentation of the results of pilot EMT-C testing.
Dr. Javed Mostafa is the Director of the Carolina Health Informatics Program (chip.unc.edu) and the Laboratory of Applied Informatics Research (http://www.informaticslab.org/). His research concentrates on information retrieval problems particularly related to search and user-system interactions in large-scale document/data repositories. He serves as the Deputy Director of the Biomedical Informatics Core at the NC Translational Clinical Sciences Institute (tracs.unc.edu) and has current research engagements in biomedical data mining, analysis, and visualization. Translating scientific advances into health care improvements is a passion for Dr. Mostafa, and based on support from the TraCS Institute he has recently co-founded a company concentrating on patient-centric decision support (keonahealth.com). At UNC, Dr. Mostafa holds a joint faculty position in information science in the School of Information and Library Science and in the Biomedical Research Imaging Center in the School of Medicine. Additionally, he holds the title of Adjunct Professor of Community and Family Medicine, School of Medicine, Duke University.
Dr. Travers is an assistant professor at the University of North Carolina at Chapel Hill (UNC-CH) School of Nursing, with a joint appointment in Emergency Medicine and adjunct status in the School of Public Health. She holds a doctorate in Information Science with a minor in Medical Informatics, and a master’s in Nursing-Health Care Systems. Dr. Travers has expertise in health care informatics service, education and research. She led the initial development of the North Carolina Disease Event Tracking and Epidemiologic Collection Tool (NC DETECT), a state public health surveillance program. Her informatics research focuses on standardization and secondary uses of emergency department (ED) data. She is currently the principal investigator of a National Library of Medicine-funded study, Adapting Natural Language Processing Tools for Biosurveillance. Dr. Travers is also a practicing emergency nurse and does clinical research in emergency department triage. She has led projects on pandemic influenza triage, and improving methods for triage of pediatric patients in the ED.