Preview

PDF, English Download (6MB) | Terms of use

Abstract

Magnetic resonance spectroscopic imaging (MRSI) has the ability to noninvasively interrogate metabolism in vivo. However, excessively long scan times have thus far prevented its adoption into routine clinical practice. Generalized autocalibrating partially parallel acquisitions (GRAPPA) is a parallel imaging technique that allows one to reduce acquisition duration and use spatial sensitivity correlations to reconstruct the unsampled data points. The coil sensitivity weights are determined implicitly via a fully-sampled autocalibration region in k-space. In this dissertation, a novel GRAPPA-based algorithm is presented for the acceleration of 1H MRSI. Autocalibration Region extending Through Time (ARTT) GRAPPA instead extracts the coil weights from a region in k-t space, allowing for undersampling along each spatial dimension. This technique, by exploiting spatial-spectral correlations present in MRSI data, allows for a more accurate determination of the coil weights and subsequent parallel imaging reconstruction. This improved reconstruction accuracy can then be traded for more aggressive undersampling and a further reduction of acquisition duration. It is shown that the ARTT GRAPPA technique allows for approximately two-fold more aggressive undersampling than the conventional technique while achieving the same reconstruction accuracy. This accelerated protocol is then applied to acquire high-resolution brain metabolite maps in less than twenty minutes in three healthy volunteers at B0 = 7 T.