Graduate Seminar Series (Dr. Oguzhan Alagoz)

Date/Time
Date(s) - January 13, 2023
10:40 am - 11:30 am

Location

Join Zoom Meeting

Categories No Categories


Oguzhan Alagoz profile picture

Oguzhan Alagoz
Date : Friday, January 13, 2023
Mode : Virtual
Affiliation: UW-Madison
Bio: Visit Page

 

Title : Optimizing Breast Cancer Diagnostic Decisions to Reduce Overdiagnosis

Abstract :  Mammography, which reduces breast cancer mortality has several negative effects such as false positives and overdiagnosis. Overdiagnosis, defined as diagnosing a cancer that would otherwise not cause symptoms or death in a patient’s lifetime, is related to the difficulty in predicting disease severity and future outcomes based on mammography. Breast cancer overdiagnosis is estimated to cost US health care system over $1.2 billion annually. Overdiagnosis rates, estimated to be around 10%-40%, may be reduced if indolent breast cancer subtypes can be identified and followed with noninvasive imaging rather than biopsy and treatment. However, there are no validated guidelines for radiologists to decide when to choose noninvasive imaging options.

We develop a large-scale finite-horizon Markov decision process (MDP) with around 4.6 million states to optimize the post-mammography diagnostic decisions. We develop and prove the optimality of a novel divide and conquer algorithm that relies on upper bounds on the optimal decision thresholds to find an exact optimal solution. To further reduce the computational burden, we project the high-dimensional MDP onto two lower-dimensional MDPs to obtain feasible and tight upper bounds for the optimal decision thresholds. We use real data from two private mammography databases to conduct numerical experiments and solve our MDP optimally. We demonstrate our model performance through a previously validated simulation model that has been used by the policy makers to set the national screening guidelines in the US. We find that our model leads to a 20% reduction in overdiagnosis on the screening population, which translates into an annual savings of approximately $300 million for the US health care system.

Bio : Oguzhan Alagoz is Proctor & Gamble Bascom Professor of Industrial and Systems Engineering at the University of Wisconsin-Madison. He is also a professor at the Department of Population Health Sciences. His research interests include stochastic optimization, medical decision making, completely and partially observable Markov decision processes, simulation, risk-prediction modeling and health technology assessment. He served as a member of ISPOR-SMDM Modeling Good Research Practices Task Force which developed recommendations for good modeling practices in state-transition modeling for the evaluation of health care decisions in 2012. He is also co-leading the University of Wisconsin Breast Cancer Simulation Model, a member of the National Cancer Institute’s Cancer Intervention and Surveillance Modeling Network, that was used to inform national breast cancer screening guidelines in the US. He is currently serving as the editor-in-chief of IISE Transactions on Healthcare Systems Engineering and associate editor for Operations Research. He previous previously served on the editorial board of Medical Decision Making and IISE Transactions. He is an elected fellow of IISE. Furthermore, he has received various awards including a CAREER award from National Science Foundation (NSF), outstanding young industrial engineer in education award from IISE, various best paper awards from INFORMS, best podium presentation award from ISPOR, and best poster award from UW Carbone Comprehensive Cancer Center. He has been the principal investigator and co-investigator on grants approximately $7 million funded by NSF and NIH.

List of Spring 2023 Graduate Seminar Series : Click Here