Vesna Prckovska1, Alard F. Roebroeck2, Pim W.L.P.M. Pullens3,4, Anna Vilanova, Bart M. ter Haar Romeny
1Biomedical Engineering, Technical University of Eindhoven, Eindhoven, Netherlands, Netherlands; 2Maastricht Brain Imaging Center, Dept. of Cognitive Neuroscience, Faculty of Psychology, Maastricht University, Maastricht, Netherlands; 3Maastricht Brain Imaging Center, Dept. of Cognitive Neuroscience, Faculty of Psychology, Maastricht University,, Maastricht, Netherlands; 4Brain Innovation B.V., Maastricht, Netherlands
High Angular Resolution Diffusion Imaging is an emerging area that overcomes the known limitations in Diffusion Tensor Imaging. One of the recent challenges in HARDI is to find acquisition schemes that can correctly represent non-Gaussian profiles in a clinical setting. This study investigates the effect of the b value and the number of gradient vector directions for Q-ball imaging and Diffusion Orientation Transform. Extensive quantitative analysis is given on synthetic software and hardware phantom data, as well as qualitative analysis on in-vivo data. Our results suggest that over a wide range of acquisition schemes DOT can outperform Q-ball.
Keywords