03/29/2023
By Mamoon Habib
The Richard A. Miner School of Computer & Information Sciences invites you to attend a master’s thesis defense by Mamoon Habib on "Estimating Causal Effects of Antidepressants on COVID-19 Outcome using Deep Learning"
Candidate Name: Mamoon Habib
Defense Date: Friday, April 7, 2023
Time: 4 to 5 p.m.
Location: Via Zoom
Thesis Title: Estimating Causal Effects of Antidepressants on COVID-19 Outcome using Deep Learning
Committee Members:
- Advisor Tingjian Ge, Computer Science Department, University of Massachusetts Lowell
- Co-AdvisorRachel Melamed, Biological Sciences Department, University of Massachusetts Lowell
- Cindy Chen, Computer Science Department, University of Massachusetts Lowell
Abstract:
The COVID-19 pandemic has caused a significant rise in hospitalization and mortality rates globally, emphasizing the need to discover effective and affordable treatments that are widely available to combat the virus and prevent severe illness. Some studies have suggested that antidepressants, commonly used to treat mental health conditions, may have the potential to reduce the severity of COVID-19. To estimate the causal effects of antidepressants on COVID-19 outcomes, we leverage large-scale observational electronic health records (EHRs) data from the National COVID Cohort Collaborative (N3C) and identify drugs that could impact COVID-19 outcomes. To adjust for a wide range of confounding variables, such as patient demographics, health conditions, and prior medications, we utilize TARNet, a deep learning model specifically designed for causal inference. Additionally, we incorporate Network Deconfounder, another deep learning model for causal inference, to account for unobserved confounding factors. Our study aims to provide valuable insights into the potential effects of antidepressants on COVID-19 outcomes, with the goal of improving patient care and treatment decisions while contributing to a better understanding of the use of deep learning approaches for causal inference on real-world data.