Dr. Jihad Obeid Research Focus
Dr. Jihad Obeid is a Professor and SmartState Endowed Chair in Biomedical Informatics in the Department of Public Health Sciences at the Medical University of South Carolina (MUSC).
His research focuses on artificial intelligence (AI), specifically, the use of deep learning models with Electronic Health Records (EHR) data and physician text-based notes for the identification of patient cohorts with specific clinical conditions and the prediction of outcomes. These models can be used in a variety of downstream tasks including, but not limited to the identification of patients for research projects, recruitment to clinical trials, population health studies, clinical decision support, and predictive modeling for health outcomes. He is well published in this field of informatics with several funded initiatives. One of his projects funded by the National Institute of Mental Health (NIMH) aims to refine an AI algorithm that analyzes text in electronic medical records to identify patients at risk of suicide. Most current models for predicting suicide risk are based on tabulated or coded data in the EHR. However, most of the information in medical records is maintained in the form of text-based clinical notes. Dr. Obeid leverages state-of-the art AI to extract information from those records to assess this risk. Once the models have been refined and tested, they could serve as an early warning system that allows busy clinicians to refer patients to appropriate and preventative mental health services. He has collaborated with other clinicians to use similar modeling methods to address other clinical conditions such as liver disease and cancer. In one of these projects this method was able to identify patients with chronic liver disease, specifically, liver cirrhosis, with very high accuracy through automated reviews of physician notes. Cirrhosis is stage of irreversible scarring of the liver in patients with chronic liver disease and was ranked the 9th leading cause of death by the Centers for Disease Control and Prevention in 2021. Prior to this technology sifting through narrative text on millions of records has essentially relied on keyword searches, where the keywords had to be provided by a clinician familiar with the disease with several rounds of trial and error. Once trained and validated, these models could lead to the identification of patients records that could be used for recruitment of clinical trials as well as for outcomes research. These studies were conducted at MUSC after an ethics review and approval by an institutional review board, which allowed the use of massive amounts of data with the utmost security measures to protect the safety and confidentiality of the data.
Dr. Obeid’s other research interests include natural language processing, social determinants of health, electronic consents, secondary use of EHR data, analysis of research networks, and biomedical ontologies. He also serves as the Associate Director of the Biomedical Informatics Center (BMIC) at MUSC and Director of the Social Determinants of Health Shared Resource, which provides informatics expertise and data integration services to promote health equity research in many areas of medicine including cancer research. Dr. Obeid holds an MD degree from the American University of Beirut, with residency training in Pediatrics at Duke University, and fellowship training in Pediatric Endocrinology at Cornell University Medical Center. However, he has a long-standing interest in computer science, which started during his undergraduate education and continued through his medical education. This interest was later formalized with Biomedical Informatics training and courses at the Division of Health. Sciences and Technology, a joint Harvard-MIT fellowship program.
In his leadership roles at MUSC, he oversees several academic and operational informatics initiatives and serves on data and AI-related consultation services for researchers. Since his arrival at MUSC in 2008, he has led the effort on multiple translational research informatics-based infrastructure projects such as, the EHR Research Data Warehouse (RDW), REDCap, and many others, and has served as principal investigator, co-investigator, and informatics leader on several federally funded projects. At the national level, he led several working groups related to translational research informatics. He is the founder and director of two graduate courses in Biomedical Informatics at MUSC and the co-founder of the Clemson-MUSC AI Hub, a community of practice in AI for healthcare applications. He serves as the co-principal investigator on the statewide project Artificial Intelligence-enabled Devices for the Advancement of Personalized and Transformative Healthcare in South Carolina (ADAPT in SC) EPSCoR Research Infrastructure Improvement Program Track-1 grant from the National Science Foundation.