Zhou Lan, PhD: Recent Statistical Methods for Diffusion MRI

Zhou Lan, PhD
Instructor in Medicine, HMS
Biostatistics Investigator, BWH

Abstract

Diffusion MRI is a popular imaging technique to investigate the tissue microstructure. The diffusion MRI data has been routinely used for various medical applications. Thus, it has become a popular research area for statisticians to develop novel and useful statistical tools. Beyond the summarized quantities (e.g., fractional anisotropy), the complicated characteristics (i.e., diffusion tensor, principal diffusion direction, orientation distribution function) become more appealing. However, these characteristics provide more informative information but pose significant statistical challenges. The key reason is that these characteristics lie on manifold other than Euclidean space; thus, we cannot apply traditional statistical tools easily. The talk will introduce the speaker’s novel methods (i.e., spatial Wishart Process, von Mises Fisher regression, Fréchet regression). Numerical results illustrate the improvements, and typical examples based on the real data introduce the usages. The talk aims to introduce these novel methods and seek applications in more medical studies based on diffusion MRI.

Short Bio

Zhou Lan is a faculty biostatistician at Brigham and Women’s Hospital, Havard Medical School (0.6 FTE with Radiology Department and 0.4 FTE with Center for Clinical Investigation). Zhou Lan earned his PhD in Statistics from North Carolina State University and received the Paige Plagge Award. Zhou has advanced knowledge of diverse statistical techniques such as spatial statistics, Bayesian computing, and longitudinal data analysis with applications to neuroimaging and clinical/epidemiological studies. His primary research focuses on developing statistical methodologies for diffusion MRI. More information on Zhou’s research.