The Open Medical Imaging Journal

2012, 6 : 80-88
Published online 2012 June 1. DOI: 10.2174/1874347101206010080
Publisher ID: TOMIJ-6-80

Automated Voxel-Wise Brain DTI Analysis of Fitness and Aging

Zhexing Liu , Mahshid Farzinfar , Laurence M. Katz , Hongtu Zhu , Casey B. Goodlett , Guido Gerig , Martin Styner and Bonita L. Marks
Department of Psychiatry, University of North Carolina, Chapel Hill, NC, USA.

ABSTRACT

Diffusion Tensor Imaging (DTI) has become a widely used MR modality to investigate white matter integrity in the brain. This paper presents the application of an automated method for voxel-wise group comparisons of DTI images in a study of fitness and aging. The automated processing method consists of 3 steps: 1) preprocessing including image format converting, image quality control, eddy-current and motion artifact correction, skull stripping and tensor image estimation, 2) study-specific unbiased DTI atlas computation via diffeomorphic fluid-based and demons deformable registration and 3) voxel-wise statistical analysis via heterogeneous linear regression and a wild bootstrap technique for correcting for multiple comparisons. Our results show that this fully automated method is suitable for voxel-wise group DTI analysis. Furthermore, in older adults, the results suggest a strong link between reduced fractional anisotropy (FA) values, fitness and aging.

Keywords:

Diffusion tensor imaging, DTI atlas, voxel-wise analysis, nonlinear warping, aging and aerobic fitness.