Samir E. Abdelrahman, MS, PhD
Assistant Professor
Department of Biomedical Informatics
University of Utah

Adjunct Associate Professor
Computer Science Department
Faculty of Computers and Information
Cairo University

Department of Biomedical Informatics
School of Medicine
University of Utah
421 Wakara, Salt Lake City
Utah 84108-3514
Office: Rm. 2034, Suite 208
Phone: (801) 587-5237
Fax: (801) 581-4297

Useful Links:
Google Scholar

Innovative computational solutions are
the enjoyable fruits of successful collaborations
among seniors, juniors, and scholars.
-- Samir E. Abdelrahman

Academic Biography

Abdelrahman earned BS in Operations Research, and MS, PhD in computer science (artificial intelligence domain). He has trained as a research scholar in highly prestigious universities; namely the University of Minnesota at Twin Cities, Vanderbilt University, the University of Illinois at Urbana-Champaign, and the University of Utah. Since 2010, he has been a research scholar in biomedical informatics domain. He has conducted several predictive analytics projects at University of Utah Healthcare to predict patient risk factors and outcomes in the areas of congestive heart failure readmissions, clinic appointment no-shows/last-minute cancellations, bleeding or readmission events following left ventricular assist device implantation, pneumonia mortality, and sepsis. Currently, Abdelrahman is a faculty at Cairo University (Egypt) and the University of Utah in computer science and biomedical informatics domains respectively.

Research Interests

Abdelrahman is interested in machine learning, natural language processing, artificial Intelligence, and visual analytics. Currently, he is focusing on integrating more than one of these techniques to develop computational solutions for healthcare applications that tackle clinical big data challanges (temporality, the curse of dimensionality, imbalance, and missingness).

Research statement

Abdelrahman 's primary goal is to analyze patient trajectories for early outcome predictions. Patient trajectories exhibit high variability from patient to patient overtime, and there is an unmet need to develop an analytic tool that handles this considerable variability. To this end, extracting and representing features from patient trajectories are the core computational techniques.

Abdelrahman 's research is threefold to develop feature extraction and representation methods that minimize the effect of various clinical data challenges for patient outcome predictions. First, he studies the similarities and dissimilarities among the trajectories over time. Second, he develops computational methods that integrate structured and free-text features. Third, he builds predictive models that depend on the first two directions. In the first direction, temporal reasoning and pattern extraction methods are investigated comparing: (i) timestamped vs. static, and (ii) fine-grain (detailed) vs. coarse-grained (summaries) patient measurements. In the second direction, information retrieval and extraction methods with natural language processing are used to handle free-text clinical notes. In the third direction, traditional statistical, classification, and deep learning techniques are integrated to develop predictive models that leverage either or both results of previous two directions.

Research Lab

Abdelrahman founded Computational Methods for Data Analytics and Predictions (CM4DAP) lab. The lab pursues his research statement goals and currently has students from the University of Utah and Cairo University. The students pursues biomedical informatics, computer science, information science, and public health MS/PhD degrees.


Copyright © 2018 Samir E. Abdelrahman. All rights reserved.