Biography

Brian Williamson, PhD, is a biostatistician with expertise in statistical epidemiology, semiparametric and nonparametric estimation theory, and high-dimensional estimation and prediction. He is interested in developing robust procedures for statistical inference when machine learning is used to address problems in public health, and in working toward accessible, affordable, high-quality healthcare for everyone. A central theme of his research is using prediction models (including machine learning and artificial intelligence) to make more accurate and efficient use of electronic health records data for research and clinical care.

Before joining Kaiser Permanente Washington Health Research Institute, Dr. Williamson completed his postdoctoral research training at the Fred Hutchinson Cancer Research Center. During his time at Fred Hutch, Dr. Williamson developed statistical methods to address issues arising in the development of biomarker panels for use in risk prediction, screening, and diagnosis. Dr. Williamson also collaborated with researchers from the Women’s Health Initiative to assess the utility of metabolomic biomarkers for predicting breast and colorectal cancer; with researchers from the HIV Vaccine Trials Network (HVTN) to aid in selecting candidate broadly neutralizing antibody regimens to advance to HIV prevention clinical trials; and was a part of the Coronavirus Prevention Network Biostatistics Team.

Dr. Williamson received his PhD in biostatistics from the University of Washington. His dissertation focused on a general framework for performing inference on model-free variable importance measures. With colleagues from the HVTN, he used this framework to identify features of the HIV viral genome that may be important in predicting viral susceptibility to the broadly neutralizing antibody VRC01.

At KPWHRI, Dr. Williamson collaborates on projects across a range of research areas including mental health, pragmatic clinical trials, and drug and vaccine safety and effectiveness.