VISE Summer Research In Progress (RiP) Series 6.22.23
VISE Summer Seminar to be led by
Nancy Newlin (CS), PhD Candidate
and
Tianyuan Yao (CS), PhD Candidate
Date: Thursday, June 22, 2023
Time: 11:45 am for lunch, noon start
Location: Stevenson Center 532
RiP Speaker #1:
Nancy Newlin, PhD Candidate, Computer Science Department
RiP Title #1:
MidRISH: Unbiased harmonization of rotationally invariant harmonics of the diffusion signal
Abstract #1:
Objective: Data harmonization is necessary for removing confounding effects in multi-site diffusion image analysis. One such harmonization method, LinearRISH, scales rotationally invariant spherical harmonic (RISH) features from one site (“target”) to the second (“reference”) to reduce confounding scanner effects. However, reference and target site designations are not arbitrary and resultant diffusion metrics (fractional anisotropy, mean diffusivity) are biased by this choice. In this work we propose MidRISH: rather than scaling reference RISH features to target RISH features, we project both sites to a mid-space. Methods: We validate MidRISH with the following experiments: harmonizing scanner differences from 37 matched healthy patients, and harmonizing acquisition and study difference on 117 matched healthy patients. Conclusion: MidRISH reduces bias of reference selection while preserving harmonization efficacy of LinearRISH Significance: Users should be cautious when performing LinearRISH harmonization. To select a reference site is to choose effect-size. Our proposed method eliminates the bias inducing site selection step.
RiP Speaker #2:
Tianyun Yao, PhD Candidate, Computer Science Department
RiP Title #2:
Diffusion MRI fiber orientation distribution function estimation using deep learning networks
Abstract #2:
Diffusion-weighted Magnetic Resonance Imaging (DW-MRI), is a pivotal imaging technique that allows for the analysis and modeling of brain tissue microarchitecture at both microscopic and millimeter scales. This technique is particularly effective for examining the structure of white matter in the brain. An essential component of dMRI is the fiber orientation distribution function (fODF), which represents the orientation and volume fraction of axon bundles within each voxel. The fODF serves as the fundamental first step for the downstream processes of tractography and connectivity analyses, providing crucial insights into the brain’s intricate network of fiber pathways. However, measurement variabilities (e.g., inter- and intra-site variability, hardware performance, and sequence design) are inevitable during the acquisition of DW-MRI. Most existing model-based methods (e.g., constrained spherical deconvolution (CSD)) and learning-based methods (e.g., deep learning (DL)) do not explicitly consider such variabilities in fODF modeling, which consequently leads to inferior performance on multi-site and/or longitudinal diffusion studies. In this talk, I will present our work that utilizes deep learning methods to explicitly reduce the scan-rescan variabilities, so as to model a more reproducible and robust brain microstructure.