AI for Actionable Healthcare: Treatment Effect Estimation on Real-world Patient Data

Date: Feb 17, 11:45am-12:45pm

Abstract: Estimating causal effects from observational data is a fundamental problem in many fields that face challenges (e.g., expensive, time-consuming, or even unethical) in running randomized control trials. Traditional causal inference methods often struggle with high-dimensional or non-tabular real-world data. In this talk, I will introduce how integrating machine learning with causal inference can advance treatment effect estimation to tackle critical healthcare challenges. First, I will present a deep learning-based propensity score weighting method that represents high-dimensional data, adjusts for confounding bias, and estimates average treatment effects for drug repurposing. Second, I will introduce a deep balancing-matching method that adjusts for time-varying confounders and
estimates individual treatment effects, supporting clinical decision-making in antibiotic stewardship for sepsis. Third, I will present foundation models trained on patient data and external biomedical knowledge graphs, designed to enhance accuracy and improve the generalizability of treatment effect estimation. To conclude, I will highlight ongoing and future work in treatment effect estimation, along with my research vision for developing and deploying actionable AI systems in real-world clinical settings.

Biographical Sketch: Ruoqi Liu is a Ph.D. candidate in the Department of Computer Science and Engineering at The Ohio State University. Her research focuses on the intersection of artificial intelligence and causal inference, with the overarching goal of enhancing accurate causal effect estimation and enabling reliable decision-making in healthcare and biomedicine. Her work has been
published in venues including Nature Machine Intelligence, ACM SIGKDD and AAAI. Her research has been featured in Nature Online, MIT Technology Review, Drug Discovery News, etc. Ruoqi is a recipient of the Presidential Fellowship from OSU, a participant in the EECS Rising Stars in 2023, and Rising Stars in Data Science in 2024.

Location and Zoom link: Zoom only at https://fsu.zoom.us/j/3195217545