Happy Thursday, Statisticians!
Apologies for the cross posting. December's Statistics in Imaging Virtual Working Group (SI VWG) will be held on 12/15 at 2:00pm EST (1:00pm CST/12:00pm MST/11:00am PST) and will spotlight two early career researchers: Dr. Yuliang Xu and Dr. Simiao Gao. Drs. Xu and Gao will be talking about Bayesian image regression with soft-thresholded conditional autoregressive priors and a Bayesian framework for tau-connectome propagation graphical modeling, respectively. All relevant details including Zoom link and abstracts are listed below.
We hope you will join us for what are sure to be two excellent talks!
Cheers,
Julia, Selena, and Yuan
--
KEY INFORMATION:
Date: December 15, 2025
Time: 2 - 3pm EST (1 - 2pm CST, 12 - 1pm MST, 11am - 12pm PST)
Zoom Link: https://arizona.zoom.us/j/82786181573
Speaker: Dr. Yuliang Xu
Title: Bayesian Image Regression with Soft-thresholded Conditional Autoregressive Prior
Abstract: In the analysis of brain functional MRI (fMRI) data using regression models, Bayesian methods are highly valued for their flexibility and ability to quantify uncertainty. However, these methods face computational challenges in high-dimensional settings typical of brain imaging, and the often pre-specified correlation structures may not accurately capture the true spatial relationships within the brain. To address these issues, we develop a general prior specifically designed for regression models with large-scale imaging data. We introduce the Soft-Thresholded Conditional AutoRegressive (ST-CAR) prior, which reduces instability to pre-fixed correlation structures and provides inclusion probabilities to account for the uncertainty in choosing active voxels in the brain. We apply the ST-CAR prior to scalar-on-image (SonI) and image-on-scalar (IonS) regression models-both critical in brain imaging studies-and develop efficient computational algorithms using variational inference (VI) and stochastic subsampling techniques. Simulation studies demonstrate that the ST-CAR prior outperforms existing methods in identifying active brain regions with complex correlation patterns, while our VI algorithms offer superior computational performance. We further validate our approach by applying the ST-CAR to working memory fMRI data from the Adolescent Brain Cognitive Development (ABCD) study, highlighting its effectiveness in practical brain imaging applications.
Speaker: Dr. Simiao Gao
Title: Tau-Connectome Propagation Graphical Modeling under Cross-sectional Neuroimaging Observations
Abstract: Tau protein spreads along brain networks in Alzheimer's disease, yet how this propagation aligns with functional connectivity remains unclear. We present a Bayesian framework that jointly models tau propagation, connectivity structure, and subgroup heterogeneity. By integrating graph-constrained infection dynamics with observed connectivity patterns, the model infers plausible propagation pathways and subgroup-specific infection sequences. Simulation results demonstrate that the framework reliably recovers infection order and captures connectivity alterations, offering a probabilistic tool for understanding tau spread and its network-level organization.
------------------------------
Julia Fisher
Assistant Research Professor
BIO5 Institute
Statistics Consulting Laboratory
By Courtesy: Department of Biomedical Engineering
University of Arizona
jmfisher@arizona.edu------------------------------