Accelerated Spatially Encoded Spectroscopy of Arbitrarily Shaped Compartments Using Prior Knowledge and Linear Algebraic Modeling

paul a bottomley,yi zhang
DOI: https://doi.org/10.1002/9780470034590.emrstm1438
2015-01-01
Emagres
Abstract:Love signal-to-noise ratios (SNIP) and long scan tunes are endemic spectroscopy. While nica g (051) offers the best SNR efficiency, the minimum scan lime is set by the need to acquire set of phase-encoding gradients steps (PEs), even if an average spectrum from a multi voxel compartment would suffice. Moreover, adding voxels to create a compartment average spectrum post acquisition sacrifices SNR compared to directly acquiring the spectrum from a correctly sized compartment. Spectroscopy with linear algebraic modeling (SLAM) is a relatively new approach to localization that directly encodes spectra in C arbitrarily prescribed compartments that are segmented from accompanying scout images. SLAM uses a tiny subset (>C) of the CSI PEs selected from central.k-space plus prior knowledge of the C01140111-riellts, 10 maximize SNR/volume and greatly reduce scan times. Spectra representing compartment averages are reconstructed by solving a set of C linear simultaneous equations that eliminate unneeded PH 110111 the standard 051 algorithm. SLAM is further optimized for SNR and COMpailfrielltal bleed errors, by pen-milting fractional- instead of integer-stepped gradients. SI AM is amenable to spatial and temporal inhomogeneily COI-MC:lions, as used routinely in cSI. Combining SIAM with sensitivity encoding (SI-NSI-) techniques from parallel MRI permits replacement of even more of the already-reduced SENSE encoding PEs, for even faster acquisitions. Adoption of the technique in initial cardiac 31P and brain 1H patient studies has demonstrated effective acceleration factors of 4- to 120-fold, SNR efficiency gains of several fold, and accurate metabolite quantification with respect to reference 051 scans.
What problem does this paper attempt to address?