Diffusion Kurtosis Magnetic Resonance Imaging and Its Application to Traumatic Brain Injury

Simple item page
dc.contributor.advisorSimon, Jonathan Zen_US
dc.contributor.advisorGullapalli, Rao Pen_US
dc.contributor.authorZhuo, Jiachenen_US
dc.contributor.departmentElectrical Engineeringen_US
dc.contributor.publisherDigital Repository at the University of Marylanden_US
dc.contributor.publisherUniversity of Maryland (College Park, Md.)en_US
dc.date.accessioned2012-02-17T06:45:33Z
dc.date.available2012-02-17T06:45:33Z
dc.date.issued2011en_US
dc.description.abstractDiffusion tensor imaging (DTI) is a popular magnetic resonance imaging technique that provides in vivo information about tissue microstructure, based on the local water diffusion environment. DTI models the diffusion displacement of water molecules in tissue as a Gaussian distribution. In this dissertation, to mimic the complex nature of water diffusion in brain tissues, a diffusion kurtosis model is used, to incorporate important non-Gaussian diffusion properties. This diffusion kurtosis imaging (DKI) is applied in an experimental traumatic brain injury in a rat model, to study whether it provides more information on microstructural changes than standard DTI. Our results indicate changes in ordinary DTI parameters, in various brain regions following injury, normalize to the baseline by the sub-acute stage. However, DKI parameters continue to show abnormalities at this sub-acute stage, as confirmed by immunohistochemical examination. Specifically, increased mean kurtosis (MK) was found to associate with increased reactive astrogliosis, a hallmark for inflammation, even in regions far removed from the injury foci. Findings suggest that monitoring changes in MK enhances the investigation of molecular and morphological changes in vivo. Extending DKI to clinical usage, however, poses several challenges: (a) long image acquisition time (~20 min) due to the augmented measurements required to fit the more complex model, (b) slow image reconstruction (~90 min) due to required nonlinear fitting and, (c) errors associated with fitting the inherently low signal-to-noise ratio (SNR) images from higher diffusion weighting. The second portion of this dissertation is devoted to developing imaging schemes and image reconstruction methods that facilitate clinical DKI applications. A fast and efficient DKI reconstruction method is developed with a reconstruction time of 2-3 seconds, with improved accuracy and reduced variability in DKI estimation over conventional methods. Further analysis of diffusion weighted imaging schemes and their affect on DKI estimation leads to the identification of two clinically practical optimal imaging schemes (needing 7-10 min) that perform comparably to traditional schemes. The effect of SNR and reconstruction methods on DKI estimation is also studied, to provide a foundation for interpreting DKI results and optimizing DKI protocols.en_US
dc.identifier.urihttp://hdl.handle.net/1903/12258
dc.subject.pqcontrolledElectrical engineeringen_US
dc.subject.pqcontrolledBiomedical engineeringen_US
dc.subject.pquncontrolledDiffusion kurtosis imagingen_US
dc.subject.pquncontrolledoptimized imaging schemesen_US
dc.subject.pquncontrolledrat brainen_US
dc.subject.pquncontrolledReactive Astrogliosisen_US
dc.subject.pquncontrolledreconstruction methoden_US
dc.subject.pquncontrolledTraumatic brain injuryen_US
dc.titleDiffusion Kurtosis Magnetic Resonance Imaging and Its Application to Traumatic Brain Injuryen_US
dc.typeDissertationen_US

Files

Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
Zhuo_umd_0117E_12701.pdf
Size:
13.54 MB
Format:
Adobe Portable Document Format
Download

Collections

UMD Theses and Dissertations
Electrical & Computer Engineering Theses and Dissertations