Lipid Suppression in CSI with Highly-Undersampled Peripheral K-Space and Spatial Priors

Berkin Bilgic¹, Borjan Gagoski¹, Elfar Adalsteinsson¹, ²

Mapping the concentration of brain metabolites using chemical shift imaging is made difficult by the presence of subcutaneous lipid signals, which contaminate the metabolites by ringing due to limited spatial resolution. Dual-density approach exploits the high-SNR property of the lipid layer to generate high-resolution lipid maps and suppress truncation artifacts. Another recent approach for lipid suppression makes use of the fact that the metabolite and the lipid spectra are approximately orthogonal, and seeks sparse metabolite spectra when projected onto lipid-basis functions. Our work combines and extends the dual-density approach and the lipid-basis penalty, while estimating the high-resolution lipid image from single-average k-space data to incur only minimal increase on the total scan time. Further, we also exploit the spectral-spatial sparsity of the lipid ring and propose to estimate it from substantially undersampled single-average in vivo data using compressed sensing, and still obtain excellent artifact suppression.