This is an old revision of the document!
Title: Leveraging longitudinal and multi-modal EHR in Survival Analysis
Abstract: Survival analysis is a fundamental statistical tool that predicts the time of an event. It has multiple healthcare applications in areas, such as hospitalization and patient mortality. Clinicians predict patient outcomes by heterogeneous modalities (e.g., text, images, and lab values). Such data poses significant challenges for traditional survival analysis techniques. In this talk, we present our effort to expand the survival analysis using multimodal, longitudinal EHR data. Our results indicate that extracted high-dimensional features from text and image provide complementary information in addition to structured EHR, and incorporating longitudinal data is useful in time-to-event prediction.
Presenter: Dr. Yifan Peng
Dr. Peng is an assistant professor at the Department of Population Health Sciences at Weill Cornell Medicine. His main research interests include BioNLP and medical image analysis, such as named entity recognition, information extraction, and eye disease diagnosis and prognosis. Before joining Cornell Medicine, Dr. Peng was a research fellow at the National Center for Biotechnology Information (NCBI), National Library of Medicine (NLM), National Institutes of Health (NIH). He obtained his Ph.D. degree from the University of Delaware. During his doctoral training, he investigated applications of machine learning in biomedical relation extraction, with a focus on deep analysis of the linguistic structures of biomedical texts.