Abstract #4451
Sparse isotropic q-space sampling distribution for Compressed Sensing in DSI
Alexandra Tobisch 1,2 , Gabriel Varela 3 , Rdiger Stirnberg 1 , Hans Knutsson 4 , Thomas Schultz 2,5 , Pablo Irarrzaval 3,6 , and Tony Stcker 1
1
German Center for Neurodegenerative Diseases
(DZNE), Bonn, Germany,
2
University
of Bonn, Bonn, Germany,
3
Biomedical
Imaging Center, Pontificia Universidad Catlica de
Chile, Santiago, Metropolitan District, Chile,
4
Linkping
University, Linkping, Sweden,
5
MPI
for Intelligent Systems, Tbingen, Germany,
6
Department
of Electrical Engineering, Pontificia Universidad
Catlica de Chile, Santiago, Metropolitan District,
Chile
The Compressed Sensing (CS) technique accelerates
Diffusion Spectrum Imaging (DSI) through sub-Nyquist
sampling in q-space and subsequent nonlinear
reconstruction of the diffusion propagator.
State-of-the-art DSI approaches that exploit CS apply
Cartesian undersampling patterns. Recently, a method was
proposed to generate 3D non-Cartesian sample
distributions that aim for isotropic sampling of
q-space. This work compares the new scheme to standard
Cartesian undersampling patterns in sparse
reconstruction of simulated diffusion signals. The
diffusion propagator and the corresponding orientation
distribution function of the reconstruction are found to
deviate less from the ground truth when using an
isotropic q-space sample distribution.
This abstract and the presentation materials are available to members only;
a login is required.
Join Here
