Sairam Geethanath1, Hyeonman Baek2,
Vikram D. Kodibagkar1,2
1Biomedical Engineering, UT
Southwestern Medical Center at Dallas, Dallas, TX, United States; 2Dept
of Radiology, UT Southwestern Medical Center at Dallas, United States
Application
of compressed sensing to 1H Chemical Shift Imaging (CSI) of in vivo human
brain data has been performed for the first time. The CSI data is sparse in
the wavelet domain along the spatial and temporal dimensions and hence can be
reconstructed with high SNR from significantly undersampled k-space. This
provides a significant reduction in acquisition time which is highly desired
for CSI. The metabolite maps generated for 3 major metabolites of
N-acetylaspartate, Creatine and Choline from 20% of the original k-space data
match closely with the corresponding metabolite maps generated for the
original k-space.