About this Event
851 Neyland Dr, Knoxville, TN 37996
https://ise.utk.edu/graduate/graduate-seminars/dr-hosseinian2-28-22/Bio:
Dr. Mohammad Hosseinian is a Postdoctoral Research Associate in the Department of Computational and Applied Mathematics at Rice University. He received his Ph.D. degree in Industrial Engineering from Texas A&M University in May 2021. He is the recipient of the 2020-2021 Outstanding Engineering Doctoral Student Award from Texas A&M University. His primary research interests are in optimization and data analytics and their applications in healthcare with focus on cancer treatment.
Title:
Optimization and Data Analytics Methods in Cancer Therapy
Abstract:
This talk focuses on applications of operations research and data science in improving cancer treatment modalities. The first part of the talk concerns optimization of drug administration in chemotherapy. The biological processes of chemotherapy take place in continuous time and are best described by ordinary differential equations (ODEs); hence, chemotherapy treatment decisions have been formulated as optimal control problems in the literature. The existing chemotherapy optimization models, however, overlook crucial discrete decisions, such as dosing and rest periods, and lack a global toxicity measure, which have limited their clinical applicability. We present a mixed-integer programming framework for chemotherapy treatment that relies on piecewise linear approximations of the physiological ODEs and, in contrast to the existing optimal control models, incorporates crucial discrete components of chemotherapy as well as an explicit toxicity measure based on chemotherapy effect on the immune system. The second part of the talk is focused on data analytics techniques to predict treatment toxicity (side effects) in radiation therapy. Radiation therapy for head and neck cancer is associated with several side effects, which may severely affect the quality of life of the surviving cancer patients. Accurate prediction of treatment toxicity is crucial to designing radiation plans with minimal side effects. We discuss shortcomings of the existing supervised-learning methods to predict side effects of head and neck cancer radiation therapy and present a semi-supervised approach to address them. We conclude the talk with directions for future work.
0 people are interested in this event