ISP AI Forum: Applying AI and Data Science to Clinical Decision-Making and Health Outcomes

Please join us for the ISP Forum, taking place from 12:30–1:30 PM in SENSQ 5317! We're excited to welcome Xingyu (Mark) Zhang as our featured speaker — his presentation details are below. Lunch will be served at noon.

Abstract

The increasing availability of electronic health records (EHRs), clinical narratives, and large-scale healthcare datasets provides new opportunities to understand clinical decision-making and improve patient outcomes. My research applies and integrates biostatistical methods, machine learning, natural language processing (NLP), causal inference, and artificial intelligence (AI) to address clinically meaningful questions using complex real-world healthcare data.

In this presentation, I will describe the evolution of my research program and highlight applications across emergency care, organ transplantation, and aging and dementia research. I will first discuss a series of studies integrating structured EHR data with clinical narratives using machine learning, NLP, transformer-based models, and large language models to predict clinical outcomes and healthcare decisions. I will then describe research examining disparities in healthcare access, utilization, and outcomes, and how this work has evolved toward causal modeling and fairness-aware AI to better understand mechanisms underlying observed disparities. I will also present applications in organ transplantation, where predictive modeling, causal inference, and simulation can inform organ utilization and allocation decisions.

Finally, I will discuss emerging work integrating multimodal healthcare data, causal methods, and AI to identify potentially modifiable determinants of health outcomes. Together, this research aims to move beyond prediction alone toward intelligent, interpretable, equitable, and clinically actionable approaches to healthcare decision-making.

Bio

Xingyu (Mark) Zhang, PhD, MS is an Associate Professor at the University of Pittsburgh with expertise in biostatistics, clinical data science, and health outcomes research. His research focuses on applying and integrating statistical methods, causal inference, machine learning, natural language processing, and artificial intelligence to analyze electronic health records and other large-scale healthcare data. His work spans clinical prediction and decision-making, healthcare disparities and health equity, organ transplantation and allocation, and aging and dementia research.

Dr. Zhang has led and collaborated on interdisciplinary research involving clinicians, biostatisticians, informaticians, and data scientists, with support from federal and institutional research programs. His current research increasingly focuses on integrating causal inference with AI and multimodal clinical data to develop approaches that are interpretable, equitable, and useful for real-world healthcare decision-making. He also mentors graduate students and trainees in quantitative methods, clinical data science, and interdisciplinary health research.