| White matter parcellation from Diffusion Magnetic Resonance Images reveals detailed 3D white matter structure and connectivity of the brain in vivo and thus is a viable tool for investigating neural differences. Measurements derived from these neural tracts or fascicles from Diffusion Tensor Images (DTI) exhibit potential as biomarkers for neurologic and neuropsychiatrie disorders.;Existing efficient white matter tract extraction approaches are based on selecting (bundling) pre-computed integration-based paths (fibers) using whole brain fiber tracking procedures. Unfortunately, tracking errors due to local noise, image artifacts (such as head motion) and numerical errors during integration-based techniques are cumulative. Prematurely terminated fibers and under-sampled fiber bundles result in incomplete reconstruction of white matter fiber tracts and hence incorrect anatomical measurements. Quantitative cross-subject tract analysis, which is critical for the detection of differences, is complicated by inefficient and inaccurate tract reconstruction and normalization from fiber bundles.;We develop and implement a parcellation method that aims for lower sensitivity to initialization and local orientation error by directly segmenting full white matter tracts, rather than reconstructing individual curves, from diffusion tensor fields. A fast, robust volumetric and intrinsically normalized solution is achieved by efficient sampling and noise- filtering using a generic parametrized tract model. Interactive tract modification schemes are also developed to resolve ambiguity in branching tracks or to perform regional investigation for long fascicles. We have packaged all of the developed methods into a multi-platform software with volumetric image and tract visualization and 82) interactive graphical user interface. Since a normalized tract can be delineated interactively in just a few seconds using our new technique, accurate tract comparisons in large samples becomes feasible. |
