Abstract #1882
3D High Resolution l1-SPIRiT Reconstruction on Gadgetron based Cloud
Hui Xue 1 , Peter Kellman 1 , Souheil Inati 2 , Thomas Sorensen 3 , and Michael Schacht Hansen 1
1
Magnetic Resonance Technology Program,
National Heart, Lung and Blood Institute, National
Institutes of Health, Bethesda, Maryland, United States,
2
National
Mental Health Institute, National Institutes of Health,
Bethesda, Maryland, United States,
3
Department
of Computer Science, Department of Clinical Medicine,
Aarhus University, Aarhus, Denmark
Applying non-linear reconstruction to high resolution 3D
MRI is challenging because of the lengthy computing time
needed for those iterative algorithms. To achieve
practical processing duration to enable clinical usage
of non-linear reconstruction, we have extended
previously published Gadgetron framework to support
distributed computing in a cloud environment. This
extension is named GT-Plus. A cloud version of 3D
l1-SPIRiT was implemented on the GT-Plus framework. We
demonstrate that a 3mins reconstruction could be
achieved for 1mm3 isotropic resolution neuro scans with
significantly improved image quality using two
dimensional acceleration factors of R=32 and 33.
This abstract and the presentation materials are available to members only;
a login is required.
Join Here
