Premium
Macroscopic orientation component analysis of brain white matter and thalamus based on diffusion tensor imaging
Author(s) -
Wakana Setsu,
NagaePoetscher Lidia M.,
Jiang Hangyi,
van Zijl Peter,
Golay Xavier,
Mori Susumu
Publication year - 2005
Publication title -
magnetic resonance in medicine
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.696
H-Index - 225
eISSN - 1522-2594
pISSN - 0740-3194
DOI - 10.1002/mrm.20386
Subject(s) - white matter , diffusion mri , fractional anisotropy , thalamus , orientation (vector space) , segmentation , artificial intelligence , neuroscience , computer science , fiber tract , pattern recognition (psychology) , independent component analysis , nuclear magnetic resonance , physics , psychology , magnetic resonance imaging , mathematics , medicine , radiology , geometry
Diffusion tensor imaging (DTI) can delineate white matter architecture based on fiber orientation. The purpose of this paper is to use the orientation information contained in DTI to study axonal organization of the brain both macroscopically and quantitatively. After performing gray/white matter segmentation using a fractional anisotropy threshold, the white matter can be further decomposed into components composed of tracts oriented along three orthogonal anatomic axes (right–left, superior–inferior, and anterior–posterior). For each component, the volume and MR parameters were quantified. To characterize the axonal architecture of the brain, this technique was applied to the entire brain using a Talairach‐based brain parcellation method and to the thalamus by manual segmentation. Reproducibility of this analysis tool was examined by repeating the measurements in the same subject, and individual differences were appreciated from the data acquired in 11 healthy volunteers. Based on the results from these preliminary data sets, this new analysis technique is expected to be an effective tool for macroscopic white matter characterization. Magn Reson Med 53:649–657, 2005. © 2005 Wiley‐Liss, Inc.