I am an Assistant Professor in the Department of Anesthesiology and the Institute for Trauma Recovery at UNC-Chapel Hill. I obtained my Ph.D. in Quantitative Psychology and M.S. in Statistics at UCLA, where I was fortunate to be advised by Dr. Peter Bentler. Before joining UNC, I worked at SAS for many years, where I developed the IRT procedure and contributed to multiple procedures in SAS/STAT software.
My research develops and applies advanced statistical and computational methods to address complex healthcare problems. A major focus of my work is improving perioperative patient care through risk prediction, treatment evaluation, and decision-support tools. A second major area centers on posttraumatic outcomes, especially chronic posttraumatic pain and related neuropsychiatric sequelae, where I work on biomarkers, subtyping, and predictive modeling. Across these areas, I am especially interested in interdisciplinary collaboration that combines clinical knowledge, rigorous analytics, and artificial intelligence to improve research and practice.
Current Direction
My recent work increasingly focuses on AI-powered research systems that connect clinical questions, longitudinal EHR data, reproducible analytics, and large language models.
Current emphasis: building research workflows with Codex and Claude that support cohort design, prompt iteration, evaluation planning, result synthesis, and manuscript development for perioperative and posttraumatic outcome studies.
Research Themes
Perioperative Care
Risk prediction and decision support
I study perioperative risk, treatment evaluation, and AI-assisted decision support tools that can help clinicians synthesize complex patient histories before surgery.
Trauma Recovery
Posttraumatic pain and neuropsychiatric outcomes
My work examines chronic posttraumatic pain, PTSD-related outcomes, biomarkers, and heterogeneous recovery trajectories using modern statistical and machine learning methods.
Methodology
Statistical modeling for complex health data
I develop and apply latent variable models, predictive methods, and computational pipelines for high-dimensional, longitudinal, and clinically rich datasets.
Recent Updates
- July 2025. Departmental funding supported our work on an AI-powered clinical decision support tool for preoperative evaluation.
- June 2025. Our paper on language biomarkers of posttraumatic outcomes was highlighted by Digital Psychiatry and Neuroscience ([YouTube video](https://www.youtube.com/watch?v=4i3QrdNy6to)).
- May 2025. Bridget Lin received the highest honor for her thesis on advanced prediction tools for posttraumatic pain and PTSD using high-dimensional methylation data.
- May 2025. Razmin Bari received the highest honor for her thesis on the causal effect of peritraumatic stress on posttraumatic pain.