Wim Van Hecke1, Alexander Leemans2, Steve De Backer1, Paul M. Parizel3, Jan Sijbers1
1University Antwerp, Antwerp, Belgium; 2CUBRIC, University of Cardiff, Cardiff, Wales, UK; 3Radiology, Antwerp University Hospital, Antwerp, Belgium
Many DTI studies are starting to use voxel based analysis (VBA) to evaluate differences in the diffusion properties between healthy diseased subjects. In this work, simulated DTI data sets with a predefined pathology are used to evaluate the sensitivity and specificity of the pathology detection comparing isotropic and anisotropic smoothing kernels with different full width at half maximum. We demonstrate that the use of anisotropic kernels can increase the sensitivity and the specificity of detecting a pathology in a VBA. We therefore suggest to apply an anisotropic smoothing method in DTI group studies to increase SNR while preserving WM boundaries.