Congratulations to PhD candidate Xinyu Zhang, alumnus Kenneth Liao (MS 2024), and associate professor Simon Vandekar on the publication of “Semiparametric confidence sets for cross-sectional and longitudinal neuroimaging” in Imaging Neuroscience last month. The peer-reviewed paper was co-authored by researchers at the University of Pennsylvania, University of Colorado Anschutz, and Vanderbilt University Medical Center’s Department of Psychiatry and Behavioral Sciences. Xinyu Zhang and Simon Vandekar are the paper’s corresponding authors. The study was funded by R01 and K24 awards from the NIH (National Institutes of Health).
The paper offers methods and software for reporting effect sizes for brain-behavior associations and including spatial confidence intervals, which assist with addressing concerns about replicability. In the words of the authors: “the tool we developed leverages recent statistical methods for confidence sets to perform inference in neuroimaging with multi-dimensional parameters and longitudinal data, expanding the research questions that can be studied using confidence sets. With the code available in the pbj R package, these methods can improve effect size reporting in neuroimaging. The statistical framework is general and can accommodate different spatial units of analysis, including voxel-wise data and atlas-based parcellations.”
Zhang graduated from China’s Southwestern University of Finance and Economics with a BS in statistics, followed by an MS in biostatistics from Yale University, where she focused on survival analysis and functional data analysis. Her previous publications include “Computationally inferred cell-type specific epigenome-wide DNA methylation analysis unveils distinct methylation patterns among immune cells for HIV infection in three cohorts” in PLOS Pathogens, for which she was lead author as well, and “Changes in functional connectivity in relapsing-remitting multiple sclerosis spinal cord measured via region-based and data-driven analyses,” another 2025 Imaging Neuroscience paper, led by chemical and physical biology doctoral student Atlee Witt and co-authored by researchers in Vanderbilt’s Institute of Imaging Science and Department of Radiology and Radiological Sciences, as well as Dr. Vandekar and senior biostatistician Trey McGonigle. Xinyu Zhang’s participation in professional conferences include organizing and chairing an invited session at SMI (Statistical Methods in Imaging) 2025 on advanced statistical methods in neuroimaging and functional data analysis; presenting an earlier version of “Semiparametric confidence sets…” at JSM (Joint Statistical Meetings) 2024; and presenting “Semiparametric confidence sets for arbitrary effect sizes in longitudinal neuroimaging” at ENAR (Eastern North American Region of the International Biometric Society) 2025.