Meeting Banner
Abstract #0498

Application of Diffusional Kurtosis to Modeling of the Cerebral Microenvironment

Edward S. Hui1, 2, Ali Tabesh1, 2, Joseph A. Helpern1, 2, Jens H. Jensen1, 2

1Center for Biomedical Imaging, Medical University of South Carolina, Charleston, SC, United States; 2Dept of Radiology and Radiological Science, Medical University of South Carolina, Charleston, SC, United States

Diffusion MRI (dMRI) has often been augmented with tissue-specific modeling in order to explicitly relate dMRI data to microstructural properties such as the sizes, orientations, volume fractions, and diffusivities of prescribed cellular compartments. One approach to tissue modeling is to exploit the close link between cytoarhitecture and the non-Gaussanity of water diffusion, which may be obtained with the dMRI technique known as diffusional kurtosis imaging (DKI). In this work, we propose a method, cerebral microenvironment modeling, which generalize the white matter model by Fieremans et al so that specific microstructural properties of the entire brain may be obtained with DKI.

Abstract PDF Presentation Video

Keywords

adult aligned allows application approximated attributable axons best biomedical biomedicine biophysical bounds brain bulk candidates cerebral challenging characterized chosen coefficient compartment compartments complexity composed computed confined considered consisting correlation defined dendrites denoted density depend dept diffusion diffusional diffusivity directional directionally distribution eigenvalues encoding entire equal exchange exchanging expected experiment explicit exploit fact fidelity formula foundation fraction full fully generalized generated gradient gray hand healthy human idealized illustrates infinitely intra intrinsic isotropic kurtosis linear links long longer mainly maps matrix measured medical metrics minimizes modeling must narrow neural note notice occupy open orientations otherwise perfectly pixels post predicted prediction predictions previously processing programs properties proportion proposed radiological radiology regression remains removing represents resolution respectively robustness rotation satisfied satisfy scanned scanner science selected since slope solution solutions solve south springer stems subcomponent subset supported system table tabulated tensor tensors termed tissue trio various vector versus viable view volume volunteer water white yields zero