Dissertation/Thesis Abstract

Tissue segmentation and classification of diffusion tensor MRI data: From models to man
by Freidlin, Raisa Z., D.Sc., The George Washington University, 2008, 152; 3304085
Abstract (Summary)

The zebrafish is an important model organism in developmental biology. Because it is optically transparent during the embryonic and early juvenile stages, it is amenable to study via powerful optical microscopy techniques. Primarily, confocal fluorescence microscopy is used to identify genes responsible for cell function, and tissue and organ formation in normal development, as well as to assess structural alterations that can be induced by “knocking out” these genes, which can sometimes be related to known diseases or developmental disorders. However, as the zebrafish approaches adulthood, it grows larger and becomes optically turbid so that these powerful optical microscopy techniques no longer work. Consequently, genetic studies in zebrafish are effectively limited to examining changes in form and function during the organism's early embryonic development. However, many have hypothesized that certain genes in zebrafish remain silent during periods of early growth, only to be expressed in the adult stages, leading to possible disease, dysfunction or dysregulation in maturity. Such studies have not been undertaken in zebrafish owing to the limitations discussed above.

Diffusion Magnetic Resonance Microscopy (MRM) provides near-microscopic resolution without relying on a sample's optical transparency for image formation. Thus, we propose using Diffusion MRM to characterize normal tissue structure in adult zebrafish, and possibly subtle anatomical or structural differences between normal and knockouts. Diffusion MRM is a noninvasive imaging technique for quantitative analysis of intrinsic features of tissues.

One aim of this work is to investigate the feasibility of using a hierarchy of models to describe diffusion tensor MR data. Parsimonious model selection criteria are used to choose among different models of diffusion within tissue. Second, based on this information, we assess whether we can perform simultaneous tissue segmentation and classification. Three hierarchical approaches are used for parsimonious model selection: the F-test t-test (F-t), proposed by Hext, the F-test F-test (F-F), adapted from Snedecor, and the Schwarz Criterion (SC). The F-t approach is more robust than the others for selecting between isotropic and general anisotropic (full tensor) models. However, due to its high sensitivity to the variance estimate and bias in sorting eigenvalues, the F-F and SC are the preferred methods for segmenting models with transverse isotropy (cylindrical symmetry). Additionally, the F-F and SC methods are easier to implement than the F-t.

A second aim is to investigate the effectiveness of unsupervised tissue segmentation and classification algorithms for DTI data.

The proposed unsupervised segmentation algorithm utilizes information about the homogeniety of the distribution of diffusion tensors within contiguous voxels. We adapt a framework proposed by Hext and Snedecor, where the null hypothesis of diffusion tensors belonging to the same distribution is determined by an F-test. Tissue type is classified according to one of four possible diffusion models, which is determined using a parsimonious model selection framework. This model selection approach, also adapted from Snedecor, chooses among different models of diffusion within a voxel using two F-tests.

Both numerical phantoms and DWI data obtained from fixed adult zebrafish and excised rat spinal cord are used to test and validate these tissue segmentation and classification approaches. The unsupervised clustering method effectively identifies distinct regions of interest in phantom and real experimental DTI data.

Indexing (document details)
Advisor: Basser, Peter J., Zara, Jason M.
Commitee: Doroslovacki, Milos, Korman, Can, Wang, Paul
School: The George Washington University
Department: Electrical Engineering
School Location: United States -- District of Columbia
Source: DAI-B 69/03, Dissertation Abstracts International
Subjects: Biomedical research
Publication Number: 3304085
ISBN: 9780549521303