Optimized workflow for in vivo small animal diffusion weighted MRI studies of white matter diseases: from acquisition to quantification.

While the number of applications of diffusion MRI has exploded in recent years, obtaining reliable and quantitative diffusion information remains a challenging task. In this project, we aim to develop diffusion weighted MRI (DWI) sequences and processing routines to obtain reliable diffusion measures within an acceptable acquisition time and at high spatial resolution to reduce partial volume effects. This would be of particular interest for in vivo pre-clinical research in small animals as mice in which the needed signal to noise ratio for reliable diffusion measures sets constraints on the spatial resolution and measure time.We will develop a diffusion –acquisition & reconstruction -workflow that reconstructs a high resolution isotropic DWI data from a set of multi-slice 2D diffusion weighted images -acquired with a 7 or 9.4 T Bruker MR scanner -with a high in-plane resolution and a lower through-plane resolution and in which the stacks of slices are differently orientated. The new reconstruction method needs to model both the different orientations of the MR images as the different orientations of the applied diffusion weighted gradients.For this super resolution at these high magnetic field, sampling the DWI with conventional fast echo planar imaging sequences will be (1) too sensitive to orientation dependent eddy current image distortions – which prevents the multi angle acquisitions and (2) suffers from local loss of signal due to B0-inhomogeneities. Therefore, we aim to develop the method based on DW-Fast Spin echo acquisition in which the images don't show B0-inhomogeneities problems and moreover can be acquired at different angles. First, we will optimize the DWI with Fast Spin Echo sampling and reconstruction. Based on this sequence, further developments will be performed to set the optimal acquisition scheme to get to super-resolution DWI: being the best combination of the set of orientations of the multi-slice stacks combined with the different directions of the DW gradients. Hereto, we can define different development steps which each will deal with specific MR acquisition and/or processing challenges : motion artifacts, multi-shot acquisition, minimization of eddy current effects, phase-wrapping, T2-modulation over k-space, denoising.The MR-sequences will be developed and implemented – in ParaVision software- on the Bruker MR scanners from the Bio-Imaging lab. The reconstruction algorithms will be developed in Matlab at the Vision lab. This new development can only be realized based on the experiences and close collaboration of both research labs.

Keywords:MRI IMAGING, RESOLUTION (SPATIAL), PROCESSING, DIFFUSION-WEIGHTED IMAGING

Disciplines:Multimedia processing, Biological system engineering, Signal processing, Medical imaging and therapy, Other paramedical sciences